Scientists Built an Enzyme That Erases a Key Driver of Aging From Human Tissue

For anyone tracking chronic inflammation, stiff joints, or age-related tissue changes, a persistent frustration has been that certain kinds of cellular damage are considered effectively permanent. A class of chemical modifications called advanced glycation end products, or AGEs, builds up on long-lived proteins throughout the body, and every existing strategy either slows their formation or mops up their chemical precursors. None can undo the damage already done. A study published in Nature Communications describes early but striking evidence that this assumption may no longer hold.

Key takeaways:

• Researchers engineered a bacterial enzyme called CMLase that chemically reverses one of the most common age-related protein modifications (CML) in human tissue.
• In lab experiments, CMLase reduced CML damage by 52% to 97% on model proteins and cut CML levels in aged human skin by more than 55%, restoring them below levels seen in younger donors.
• This is early, in-vitro research; the enzyme has never been tested in a living person, and fundamental questions about safety, delivery, and whether reversal translates to functional tissue recovery remain unanswered.
• The platform that produced CMLase can potentially be adapted to target other forms of protein damage linked to aging and diabetes.

CMLase Reverses Nε-Carboxymethyl-Lysine Damage on Aged Human Proteins

The study centers on a specific AGE called Nε-carboxymethyl-lysine, or CML: a stable chemical tag that forms when sugars or oxidized fats react with lysine residues (the building blocks of proteins), locking proteins into abnormal configurations. CML accumulates on collagen in blood vessel walls, on crystallins in the eye lens, and on proteins throughout the skin. The researchers engineered a new enzyme, which they named CMLase, specifically to cleave this modification and release the lysine residue back to something close to its original state. In laboratory conditions, CMLase reduced CML signals by 52% to 97% across a range of proteins, including collagen, casein, and hemoglobin. On human lens proteins, it cut total CML concentration by 45% as measured by mass spectrometry and by 78% as measured by a separate antibody-based assay. Treated samples from the arterial tissue of a 75-year-old donor showed more than a 70% reduction in CML staining. Elderly skin treated with CMLase dropped to CML levels below those measured in a 31-year-old donor, a before-and-after contrast that is arresting even as a lab result.

Why AGE Accumulation Matters for Inflammation, Tissue Stiffness, and Disease

CML and related AGEs are not passive bystanders. They crosslink collagen fibers, which is one reason arteries and skin stiffen with age. They also activate a receptor called RAGE (receptor for advanced glycation end products), which triggers the inflammatory NF-κB pathway and releases pro-inflammatory cytokines and profibrotic growth factors. In the brain, CML accumulation has been linked to oxidative stress and mitochondrial damage in microglia, the brain's resident immune cells, disrupting the homeostatic environment neurons depend on. For people living with diabetes, AGE formation accelerates dramatically because high blood glucose provides more raw material for the reactions that produce CML. Conditions like chronic inflammatory disease and vascular stiffening that worsen with both aging and diabetes are plausibly downstream of this chemistry. The CML-RAGE signaling axis is also relevant to anyone tracking IBD, skin aging, or cardiovascular risk, because chronic low-grade RAGE activation appears to be a shared thread across those conditions. Understanding how the gut microbiome and systemic inflammation interact has reshaped how researchers think about AGE-driven tissue pathology; dysbiosis can independently raise circulating AGE precursors, adding another layer of complexity.

How the Team Engineered CMLase Through Directed Evolution

The starting point was a FAD-dependent enzyme called glycine oxidase (GO) from the bacterium Bacillus subtilis, chosen because CML's chemical structure resembles glycine. That starting enzyme could weakly process free CML in solution but showed no detectable activity on CML embedded inside a protein chain, which is where the relevant damage actually sits. The team identified a related enzyme from Calidithermus roseus (CrGO) that had a 20-amino acid deletion near its active site, opening a cavity large enough to accommodate a CML residue within a peptide. Using a genetic selection system in which E. coli bacteria with a lysine deficiency could only survive and grow if they could extract lysine from CML (by oxidizing the CML modification off a peptide substrate), the researchers ran five rounds of directed evolution, introducing random mutations through error-prone PCR and building targeted libraries around the enzyme's active-site loops. The final variant, CrGO-897, now called CMLase, carries 15 amino acid substitutions and a two-amino acid deletion relative to the starting CrGO scaffold. It achieved a greater than 10-fold improvement in catalytic efficiency on peptide substrates. Proteomics confirmed that CMLase reduced CML at 30 out of 33 modified lysine sites on albumin (BSA), with 21 of those sites showing more than 50% reduction, giving a site-by-site map of where the enzyme works and where it doesn't.

What This Could Mean for People Managing Age-Related or Inflammatory Conditions

For someone dealing with tissue-level consequences of aging or diabetes, including cardiovascular stiffness, kidney disease, retinal changes, or chronic skin inflammation, the appeal of a therapy that could reverse existing protein damage rather than just slow new formation is obvious. The practical translation remains distant. CMLase has only been tested on tissue samples in a dish; it has never entered a living organism. Researchers do not yet know whether the enzyme can reach dense connective tissue in a functioning organ, how the immune system would respond to a bacterial-derived protein, or whether chemically removing CML modifications from aged collagen would actually restore elasticity or quiet RAGE signaling in vivo. The byproducts of CMLase's activity are hydrogen peroxide and glyoxylate; the team notes these should be negligible at the low CML concentrations found in tissue and given the enzyme's specificity, but that has not been confirmed in a living system. People with IBD, IBS, eczema, or other conditions in which chronic RAGE-driven inflammation is suspected as a contributing factor should know this research exists and is genuinely novel, while also recognizing that no clinical intervention follows from it yet. Ongoing challenges in developing microbiome and tissue-targeted therapeutics illustrate just how long the road from a promising enzyme to a deliverable treatment typically runs.

Unanswered Questions: Immunogenicity, Delivery, and What CML Reversal Actually Fixes

The researchers are candid about what the study does not establish. Functional restoration was not measured: no one tested whether treated tissue became more elastic or less inflammatory. Tissue penetration into intact, living organs has not been evaluated. Because CMLase is derived from a bacterial enzyme, it will likely trigger an immune response if injected into a human, requiring either de-immunization (redesigning surface residues to be less recognizable) or some form of immunomodulation. The study also focused only on CML; aging tissue accumulates dozens of chemically distinct AGEs, some of which, like glucosepane, form irreversible crosslinks that CMLase cannot touch. The authors note that the directed-evolution platform they developed could in principle be pointed at other AGEs, including glucosepane, which would represent a significant expansion of the approach.

What Comes Next for CMLase Research

The immediate next steps involve pharmacokinetics and immunogenicity studies in animal models, followed by work to understand whether CML removal in living tissue translates to measurable functional benefit. The directed-evolution selection platform is also being positioned as a generalizable tool: the same genetic screening approach that produced CMLase could be adapted to evolve enzymes targeting other pathogenic AGEs. Given how deeply the gut microbiome shapes metabolic and immune function in ways that directly influence AGE precursor levels, researchers may also want to explore whether microbiome-level interventions could complement enzymatic AGE reversal as part of a broader strategy. For now, CMLase represents a proof-of-concept that stable, long-assumed-irreversible protein modifications can be enzymatically reversed in human tissue. That is a meaningful shift in what the science considers possible, even if the path to a therapy is long.

Common Questions

What is CML and why does it accumulate with age? CML (Nε-carboxymethyl-lysine) is a chemical modification that forms when sugars or oxidized fats react with lysine residues in proteins. It builds up over a lifetime on long-lived proteins like collagen because the body has no natural enzyme to remove it, only systems that intercept its chemical precursors before they can react.

Is CMLase something I can take now? No. CMLase exists only as a laboratory reagent at this stage. It has been tested on tissue samples outside the body; it has never been administered to an animal or human, and significant safety, delivery, and efficacy questions must be resolved before any clinical use is possible.

Does this research apply to diabetic tissue damage? Potentially, yes. AGE formation accelerates in people with diabetes because elevated blood glucose dramatically increases the supply of chemical precursors. CMLase reduced CML in multiple protein types relevant to diabetic complications (collagen, hemoglobin, retinal proteins), but these experiments were conducted in vitro and no diabetic animal or human data exist yet.

Could gut health affect AGE accumulation? There is a plausible connection. A disrupted gut microbiome can increase systemic inflammation and alter metabolism in ways that raise circulating AGE precursors, though the direct link between microbiome composition and CML accumulation in tissue has not been established in clinical research.

References

1. Narisa Trabosh, Jason Smith, Maggie Yun-Hsuan Hsu, et al. (2026). Reversal of protein chemical aging by enzymatic deglycation. Nature Communications. https://doi.org/10.1038/s41467-026-75141-2
2. Yu-Jie Zhang, Sha Li, Ren-You Gan, Tong Zhou, Dong-Ping Xu, Hua-Bin Li (2015). Impacts of Gut Bacteria on Human Health and Diseases. International Journal of Molecular Sciences. https://doi.org/10.3390/ijms16047493
3. The Human Gut Microbiome in Health and Disease. NIH. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7910902/
4. Katie Wright (2025). The Future of Microbiome-Based Therapeutics. European Medical Journal. https://doi.org/10.33590/emj/ctgy2726
5. The gut microbiome in health and in disease. NIH. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5031175/