Enzymatic protein repair proposed as a third option alongside degradation and replacement in a review by Lyakhovich and Ferrucci

Enzymatic protein repair proposed as a third option alongside degradation and replacement in a review by Lyakhovich and Ferrucci

On October 8, Ageing Research Reviews published a review by Alex Lyakhovich of Sabanci University and Luigi Ferrucci, scientific director of the US National Institute on Aging. The authors matched types of protein damage with the enzymes that repair them.

Cells usually break down a damaged protein and make a new one. In protein repair, an enzyme corrects chemical damage to the protein itself. The review asks:

“Why do cells have complex and highly redundant mechanisms for repairing nucleic acids, yet only minimal, narrowly specialized systems for directly repairing proteins?”

The conventional reasoning is that proteins carry no hereditary information, so tolerating or removing damaged proteins is simpler than repairing them. The authors explain that repair enzymes are scarce because each type of chemical damage requires a specific enzyme, and cells must distinguish damage from the modifications they use to regulate proteins.

The authors estimate that breaking down and rebuilding a protein of 317 amino acids costs a cell 400 to 430 ATP molecules if 70% of the amino acids are reused. Enzymatic repair may require 1 to 10 ATP molecules when the enzyme is already available and the damage is minor; for severe damage, replacing the protein is more reliable.

Some proteins undergo almost no replacement. The lens is enclosed in a capsule and has no blood vessels, so the crystallins in its nucleus, formed almost entirely around birth, last a lifetime. A 2000 estimate put the time needed to replace half the collagen molecules at 117 years in articular cartilage and 15 years in skin. The authors suggest that protein replacement may carry risks in neurons within established memory circuits. In their view, a repair enzyme could provide an evolutionary advantage where protein replacement is difficult.

For example, the repair enzymes MsrA and MsrB remove oxygen added to the amino acid methionine; their activity declines with age in many tissues. Aspartate residues in proteins spontaneously convert to isoaspartate, which can impair protein function; the enzyme PCMT1 initiates the conversion back to aspartate. Mice lacking this enzyme died from seizures at an average age of 42 days, whereas restricting its activity to neurons increased lifespan fivefold. SIRT3 was considered a metabolic regulator, but a 2015 analysis in mice indicated that its main role is to remove acetyl modifications that arise spontaneously, a form of protein repair.

Carboxymethyllysine (CML), a product of chemical damage to the amino acid lysine, accumulates with age in collagen and lens proteins. The bacterial enzyme MnmC converted CML back to lysine in vitro, while the review reports that no related enzyme with this activity has been found in eukaryotes. Revel Pharmaceuticals and collaborators engineered another bacterial enzyme, glycine oxidase, into CMLase: treatment of sections of donor aorta and skin reduced the detected CML signal.

In fruit flies, overexpression of PCMT increased average lifespan by 32 to 39% at 29 °C, with no lifespan extension at 25 °C; the study authors suggest that the enzyme may become limiting under mild stress that accelerates protein damage. In female mice overexpressing MsrA in the cytoplasm, the risk of death rose more slowly with age, while lifespan remained unchanged.

To determine whether repair falls behind the accumulation of damage with age, the authors propose measuring repair rates and mapping damage and repair across tissues and ages. They suggest that repairing collagen and crystallins may require targeted enzyme delivery, and that genes of unknown function may encode repair systems yet to be discovered.