Research out of Stanford Medicine and published in Eye has identified a potential way to measure ocular aging via an artificial intelligence (AI) approach.
Give me some background.
Per the study authors, analyzing single cells is key to understanding disease mechanisms in humans; however, analyzing cells in the brain and eye “is impractical,” as any tissue biopsies would lead to serious damage in those non-regenerative organs.
Thus, that brings us to this research: the authors sought to avoid this damage by integrating “proteomics of liquid biopsies with single-cell transcriptomics from all known ocular cell types,” in order to identify the cellular origin of proteins present in the eye.
To note, “proteomics” is the large-scale study of proteins and their cellular activities, while “transcriptomics" is a cell or tissue’s complete set of all RNA molecules.
And the goal?
The goal: to accurately predict a healthy person’s age based on an “eye-aging” clock.
Exactly how did they do this?
To start with, the researchers collected aqueous fluid from 46 healthy patients with the following ocular diseases:
- Diabetic retinopathy (DR)
- Retinitis pigmentosa (RP)
- Uveitis
They then used a technique they developed for studying eye fluid (prior to this current study) to analyze 5,953 proteins in the fluid.
Of that number, 26 were identified as proteins that could be used to predict aging in the eye.
And how does AI come into play?
The researchers trained an AI algorithm to develop an eye-aging clock—an approach called tracing expression of multiple protein origins (TEMPO)—that generates a “proteomic clock” to predict a person’s age based on their protein profile.
Explain this technique.
TEMPO works by tracing proteins to a specific type of cell where RNA (which creates the proteins) is located.
In essence, it gives researchers a look at and better understanding of the cell-based origin of these disease-targeting proteins.
Gotcha. So what were the findings?
Patients with diseased eyes contained proteins that indicated a higher age:
- Early-stage DR = 12 years older
- Late-stage DR = 31 years older
- RP = 16 years older
- Uveitis = 29 years older
Were the actual cell types the same?
Interestingly enough, the cells associated with increased age for each disease were different:
- Late-stage DR = vascular cells
- RP = retinal cells
- Uveitis = immune cells
Anything else surprising?
Indeed there was …
Cells often targeted in common ocular disease treatments may not actually be the cells that are involved (which means today’s standard therapies might need to be reexamined).
Case in point: Diabetes drugs target blood vessel cells; however, it turns out that the largest protein increases during DR progression where in immune cells (macrophages).
Further, accelerated aging was present in some cells before disease symptoms even began—which could be a major breakthrough in treating early and preventing disease damage.
So what does all this mean?
According to investigator Vinit Mahajan, MD, PhD, a surgeon and professor of ophthalmology at Stanford University, targeting both aging and disease cells could make treatment more effective, as the two appear to act separately but simultaneously to damage the eye.
And for the future?
Dr. Manhajan also indicated that the TEMPO technique and eye-aging clock may be used in other non-ocular organ fluids, like joint fluid or liver bile.
Additionally, this technique could potentially aid in running more successful clinical trials—identifying the cells that drive a specific disease and aging could result in better accuracy.