A large plasma proteomics study shows that aging patterns in specific cell types may help identify who is more vulnerable to disease and who is more resilient.
Study: Plasma proteomic signatures of cellular aging predict human disease. Image Credit: Kateryna Kon / Shutterstock
A recent study published in the journal Nature Medicine suggests that estimating cell-type-specific aging signatures from blood-based plasma proteins could help scientists predict disease risk.
Analyzing over 7,000 plasma proteins in over 60,000 individuals, they found that accelerated aging signatures in specific cell types were associated with a higher risk of disease. For instance, extreme astrocyte aging increased the risk of incident Alzheimer’s disease (AD) among people with the APOE4 genotype. On the other hand, young nerve and immune cells seemed to have a protective effect, improving survival outcomes. These findings suggest that protein profiling could one day help researchers stratify risk and explore more personalized approaches to prevention and treatment.
As people age, body cells undergo several changes that may alter structure and function. These changes could also predispose individuals to chronic diseases. Scientists are now developing strategies that could enable early identification of at-risk individuals. If detected early, patients could receive timely treatment before the disease progresses to advanced stages. Such proactive approaches may improve disease prognosis and ultimately enhance overall well-being.
By exploring the biological mechanisms that drive aging, researchers may identify molecular targets for interventions that could reduce the risk of age-related diseases. Cellular-resolution epigenetic and transcriptomic clocks can measure cellular aging by analyzing gene activity and nucleic acid modifications, but often require tissue biopsies, laboratory samples, or animal experiments.
About the Study
In the present study, researchers investigated whether plasma protein analysis could be used to estimate aged cell populations linked to disease. To do so, they analyzed more than 7,000 proteins in plasma samples of 60,542 individuals using machine learning models. They then linked cell types to their corresponding plasma proteins using the Human Protein Atlas. The analysis helped researchers estimate the biological age of over 40 cell types across the nervous, immune, endocrine, and musculoskeletal systems. They also estimated mortality risks over 15 years of follow-up.
To confirm the reliability of the findings, the researchers used aging clocks derived from two separate plasma protein profiling platforms. The SomaScan platform measured 7,289 proteins, whereas Olink assessed 2,923 protein markers. They subsequently evaluated these models across three large population cohorts: the Global Neurodegeneration Proteomics Consortium (GNPC, 14,281 participants), the 1946 National Survey of Health and Development (NSHD, 1,803 participants), and the United Kingdom Biobank (UKB, 44,458 participants).
The researchers used the Knight Alzheimer’s Disease Research Center (Alzheimer's Disease Research Center">KADRC) and UKB data to train models for SomaScan and Olink analysis, respectively. After quality assessment, 43 SomaScan cell-type models and 48 Olink models were retained for downstream analysis. They also developed a polycellular aging risk score (PARS) to classify mortality risk based on cellular aging determined using these datasets and proteomic platforms.
The team calculated ‘age gaps’ for individual cell types based on the difference between the cell-specific predicted biological age and the model-estimated biological age expected for an average individual of the same chronological age. They then assessed whether these age gaps correlated with disease status. They also compared cell type-specific biological age with the Clinical Dementia Rating (CDR) and Preclinical Alzheimer Cognitive Composite (PACC) scores to identify cell types with the strongest relationships with dementia severity and cognitive decline.
Results
The researchers found that cell type-specific aging signatures were associated with incident disease risk. Cellular aging also predicted mortality risk over the 15-year follow-up period. Compared to apolipoprotein E3 (APOE3) carriers, people with the APOE4 genotype exhibited older astrocytes and younger macrophages. The presence of extremely aged astrocytes tripled the risk of incident AD among participants with two APOE4 alleles. Young astrocyte populations, in contrast, reduced disease risk.
AD was associated with accelerated aging in several cell types throughout the body, including brain cells involved in nerve insulation and support, intestinal lining cells, and pancreatic cells involved in metabolism and insulin production. Accelerated aging of oligodendrocyte precursor cells and inhibitory neurons showed the strongest associations with CDR scores. The team observed similar associations using blood pTau-217 levels and PACC scores, particularly for oligodendrocyte precursor cells.
The researchers observed similar findings related to the musculoskeletal and respiratory systems. Individuals with aged skeletal myocytes were 12.7 times more likely to develop amyotrophic lateral sclerosis (ALS) than those with young skeletal myocytes. Among current smokers, extreme aging in both alveolar type 2 cells and the broader respiratory epithelial lineage increased the risk of lung cancer by 58% compared with current smoking alone.
Cellular aging also affected survival outcomes, with young immune and nerve cells associated with better survival. Individuals with normal cellular aging had about a 90% survival rate over 15 years. People with more than 20 extremely aged cell types had much lower survival rates, approximately 34%.
However, the authors noted that the findings require validation in broader populations, as the models relied on Human Protein Atlas cell-type annotations, plasma proteins may not always directly reflect cellular gene activity, and the study cohorts were predominantly older and Caucasian.
Conclusions
The findings suggest that plasma proteomic signatures of cellular aging could help scientists determine disease susceptibility and survival outcomes. This could help researchers refine future risk stratification, study disease mechanisms, and identify high-risk groups for further monitoring or research.
Based on the findings, strategies that can prevent or halt cellular aging could potentially help lower disease burden and improve longevity. If confirmed across broader, more diverse populations, clinicians could eventually incorporate protein profiling tests into future disease risk stratification or targeted monitoring strategies to enable risk stratification and ultimately improve the standard of care for affected individuals.
Download your PDF copy by clicking here.
Frailty flags poorer blood pressure control in older women