A Catalan AI detects the traces of aging in blood stem cells

BarcelonaAging is a natural process and not a disease. Even so, accumulating years is considered the main risk factor for the most prevalent health problems on the planet, such as cancer or Alzheimer's. The passage of time leaves traces inside cells and causes them to deteriorate. And now, Catalan researchers, led by the Bellvitge Biomedical Research Institute (Idibell), have managed for the first time to identify these traces in blood stem cells in microscopic images.

In a study conducted with mice, they have developed a deep learning artificial intelligence tool capable of distinguishing between young stem cells and aged stem cells by analyzing three-dimensional images of the nucleus and detecting subtle age-associated changes in how DNA is stored.

The new tool, which they have named ChromAgeNet, opens the door to exploring possible strategies to preserve or restore the function of these cells. “The tool we have developed is important because it helps us understand the mechanisms that can be a therapeutic target,” points out Maria Carolina Florian, Icrea researcher at the Idibell regenerative medicine program.

A library in a shoebox

The work, in which researchers from the Barcelona Institute for Global Health and the Barcelona Supercomputing Center (BSC-CNS) also participated, has focused on the hematopoietic system, the set of organs and tissues responsible for producing blood cells, including immune system cells. Recent studies have revealed that, precisely, this defense system plays a key role in the aging process of the entire organism.

It is known that aging modifies the way the cell organizes genetic material within the nucleus. This is because DNA is like a library made up of millions of books compressed into a shoebox, which would be the cell's nucleus. The way these books—the genes—are placed determines which ones can be consulted and, therefore, can function, and which ones remain out of reach or turned off.

To see if these modifications could function as a kind of age clock, the researchers obtained high-resolution 3D images of the nuclei of hematopoietic cells and stained them with a marker that allowed them to visualize the DNA. Then, they trained ChromAgeNet with these images so that it could distinguish between young and aged stem cells based on the 3D structure of the nucleus. And they discovered that the aged ones had a "more disorganized" structure, in Florian's words, and this affected their function.

Drugs to reverse the effects of aging

In the study –which is published in the journal Aging Celland which has counted on European funding through an ERC Consolidator grant, from the Ministry of Science and the La Caixa Foundation– Florian and his team have treated aged cells with compounds that modify the way DNA is packaged inside the nucleus, in a structure called chromatin. With some of these compounds, the organization of the chromatin acquired an appearance more similar to that of young cells, a change that ChromAgeNet was able to detect. The result suggests that these treatments can modify some of the changes in chromatin associated with aging, but it still remains to be verified if this translates into a recovery of cell function.

The next step will be to verify if this aging signature can also be identified in human cells and if the tool can serve to find treatments capable of modifying it.