If this research succeeds, it could change the way we think about aging: has artificial intelligence just solved the problem of aging?
In recent hours, everyone is talking about the possibility of restoring a younger appearance to the skin. But the real story lies elsewhere entirely.
A study published a few days ago in Nature Communications, in collaboration with Revel Pharmaceuticals, Calico, and the University of Colorado, presents a new approach to removing one type of molecular damage previously considered irreversible.
At the center of the research is a molecule called CML, one of the end products of glycation (AGEs). This is damage that accumulates over decades when sugars react with proteins in the body.
The process is similar to bread browning in an oven: slow, cumulative, and almost irreversible.
The problem is that CML does not simply remain in the body. It binds to collagen and elastin in various tissues: in the skin, it impairs elasticity; in the eye lens, it accumulates over the years; and especially in artery walls, where it can contribute to blood vessel stiffening and increased chronic inflammation via the RAGE receptor.
For over four decades, it was assumed that the human body had no way to break down this molecule.
All treatments studied to date primarily attempted to slow the formation of new damage, not to remove what already existed.
This is where the breakthrough comes in.
The researchers designed a completely new enzyme that does not exist in nature.
They combined artificial intelligence tools, including AlphaFold, scanned approximately 45,000 protein structures, and then performed five rounds of directed evolution on over 500 million different variants, until they developed the CMLase enzyme.
Its goal is simple yet ambitious: to identify the CML molecule, remove it from the protein, and leave the protein in a state closer to its original form.
The laboratory results were impressive: in aortic tissue from approximately 75-year-old donors, over 70% of CML molecules were removed. In human skin from elderly donors, over 55% were removed, to levels lower than those measured in the skin of a 31-year-old. In eye lenses, a reduction of 45%–78% was recorded.
The researchers noted that they initially expected only about 20% removal, and therefore the results surprised them as well.
If it turns out in the future that CML removal also improves tissue function, the implications could be much broader than aesthetics.
Arterial stiffness is one of the main causes of systolic hypertension, heart failure, and stroke.
Currently, there are treatments that slow the process, but no treatment that removes damage that has already accumulated; if this damage can also be reversed in a living body, this represents a potential for significant change in aging medicine.
However, it is important to maintain perspective.
The research was conducted on donated tissues tested outside the body, and not in living humans.
It has not yet been proven that CML removal actually restores arterial flexibility or improves their function. Additionally, delivering a large enzyme into deep tissues in the body is a significant biological and engineering challenge, and clinical trials are still far off.
Investor enthusiasm is already palpable.
On the All-In podcast, David Friedberg and Chamath Palihapitiya argued that if the enzyme could be turned into a cream or a simple treatment, it represents a potential multi-trillion dollar market.
They may be right about the economic potential. But the true value of the research is not a cosmetic cream.
It lies in the fact that for the first time, a type of molecular damage, considered irreversible for decades, was successfully removed from human tissue.
If this approach proves effective in a living body as well, it may not be just another treatment that slows aging, but the beginning of an era where some of the cumulative damage of age becomes repairable.
Artificial intelligence is changing the world, and what's amazing is that, apparently, we haven't seen anything yet.