AI identifies key gene driving aging in blood stem cells

Stem cells possess the unique ability to self-renew and transform into different types of cells. For example, rare hematopoietic stem cells found in bone marrow can become any type of blood cell. Throughout life, they can produce new blood cells to replace those lost because of infection, bleeding, or chemotherapy.

However, as people and animals age, hematopoietic stem cells become less effective. Their numbers increase, but their ability to rebuild the blood system declines. These changes may contribute to anemia, reduced immunity, blood clots, and age-related blood disorders. Researchers at Tohoku University used AI to predict which genes may lead to aging in stem cells, and subsequently tested these candidates in animal studies. This research provides a clearer molecular framework for understanding why blood production becomes unbalanced with age, shining light on an issue that affects everyone as they grow older.

The researchers analyzed individual hematopoietic stem cells from mice of different ages. They found that aged stem cells simultaneously activate two gene programs: one that preserves a highly immature stem-cell state and another that prepares the cells to produce platelets. These changes began gradually, with the immature program increasing before birth and platelet-related genes increasing after birth. This suggests that stem-cell aging develops continuously rather than appearing suddenly in old age.

To identify the genes controlling this process, the team used Geneformer, an AI model trained on gene-expression data from about 30 million cells. They further trained the model using approximately 160,000 young and aged blood stem and progenitor cells. The AI predicted which genes could shift young stem cells toward an aged state.

AI helped us find 143 promising candidates, which we then screened in the lab and narrowed down to a key control point of aging: a gene-regulating factor called Pbx1."

Keiyo Takubo, Tohoku University

In aged stem cells, Pbx1 was strongly connected to genes involved in stem-cell immaturity, aging, and platelet production. Increasing Pbx1 in young stem cells reproduced many features of aged cells. Up to 73.3% of the genes activated by Pbx1 were also increased in aged stem cells. After transplantation into mice, these cells produced fewer red blood cells and showed a relative increase in platelet production. They proposed that part of the reason for this reduced red blood cell development was Pbx suppressing another gene called Gata1.

"Aged blood stem cells are often described simply as cells that have lost function," says Takubo. "But our findings show that they aren't a weaker version. They just enter a different, stable state with their own tendencies."

These findings - combining AI prediction, large-scale screening, multi-omics analysis, and transplantation experiments - reveal key points of one of the pathways that leads to aging in hematopoietic stem cells. Future studies will determine whether the same mechanism operates in humans and whether it contributes to anemia, thrombosis, clonal hematopoiesis, or blood cancers.

The findings were published in Science Advances on August 22, 2026.

 

Source:
Journal reference:

Kobayashi, H., et al. (2026). Geneformer-guided multiomics integration identifies Pbx1 as a network hub of hematopoietic stem cell aging. Science Advances. DOI: 10.1126/sciadv.aeb1346. https://www.science.org/doi/10.1126/sciadv.aeb1346

Comments

The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of News Medical.
Post a new comment
Post

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.

You might also like...
KAIST research opens new path toward releasing 'molecular lock' and reversing cell's fate