Spatial aging clocks reveal how T cells and neural stem cells shape brain aging

By Vijay Kumar MalesuReviewed by Susha Cheriyedath, M.Sc.Dec 22 2024

Discover how cutting-edge spatial aging clocks decode the aging brain, uncovering the dual roles of immune and stem cells in reshaping our understanding of cognitive decline and rejuvenation.

Study: Spatial transcriptomic clocks reveal cell proximity effects in brain ageing. Image Credit: VectorMine / Shutterstock

In a recent study published in the journal Nature, researchers developed spatial aging clocks using single-cell transcriptomics to explore cell-type-specific interactions and their impact on brain aging, rejuvenation, and disease.

Background

Brain aging significantly increases the risk of neurodegenerative diseases like Alzheimer’s (a progressive brain disease causing memory loss) and dementia (a decline in cognitive abilities). While previous research has explored molecular changes in the aging brain at single-cell resolution, these studies lack spatial context, particularly at scale. Without a systematic understanding of spatiotemporal changes, including local cell neighborhoods and cell-cell interactions, crucial insights are missed. High-throughput spatial omics show promise for advancing this understanding, but current studies fail to capture both spatial and temporal resolution at the single-cell level, especially in geriatric ages when cognitive decline is most apparent. This study addresses these gaps by introducing spatial aging clocks, which offer a new computational framework to predict cell-specific aging and explore cell proximity effects. Further research is needed to develop advanced computational tools to analyze these spatial interactions.

About the Study

In the present study, male C57BL/6JN mice were used for the aging and exercise cohorts, while male whole-body inducible OSKM (POU class 5 homeobox 1 (Oct4), SRY (sex determining region Y)-box 2 (Sox2), Kruppel-like factor 4 (Klf4), and Myelocytomatosis oncogene (c-Myc)) mice were used for the partial reprogramming experiment. Mice were housed in groups under standard conditions, with at least three weeks of acclimatization prior to experiments. The aging cohorts included mice of varying ages, ranging from 3 to 34 months, with coronal and sagittal brain sections collected for transcriptomic analysis. The exercise experiment included young and old sedentary and exercise mice, while the partial reprogramming experiment used young and old OSKM mice with doxycycline treatment. All animal procedures were approved by the Stanford University Institutional Animal Care and Use Committee (IACUC) and the Veterans Affairs Palo Alto Committee on Animal Research.

For sample collection, mice were euthanized, and brains were snap-frozen in an Optimal Cutting Temperature (OCT) compound. Ribonucleic acid (RNA) sequencing data was obtained using the Multiplexed Error-Robust Fluorescence In Situ Hybridization (MERFISH) platform with a custom 300-gene panel. The panel included markers for various cell types and aging-related genes. Brain sections were processed for MERFISH with tissue permeabilization, hybridization, and imaging following the Vizgen protocol. After image collection, cell segmentation and transcript allocation were performed using Cellpose. Data were preprocessed by filtering out low-quality cells, and gene expression normalization was applied.

Machine learning models were trained on the transcriptomic data for spatial aging clocks to predict age based on spatial gene expression patterns. The proximity effects of T cells and neural stem cells on neighboring cells were analyzed by comparing transcriptomic changes in nearby and distant cells. Statistical analyses included Pearson correlation and Mann-Whitney U-test, with visualization performed using various plotting tools.

Study Results

A spatial transcriptomics atlas of the aging mouse brain was created to map gene expression across the entire lifespan. The dataset encompassed 2.3 million high-quality cells from different brain regions, spanning ages from 3.4 to 34.5 months. The MERFISH method identified 18 cell types, including neurons, glial cells, and immune cells, and showed how these cells localized to their respective regions.

The study revealed significant changes in cell proportions with age. For example, microglia and T cells increased with age, while neural stem cells (NSCs) and oligodendrocyte progenitor cells (OPCs) decreased. T cells showed a substantial increase in numbers across all regions, while NSCs were mainly found in the neurogenic niche and decreased over time. These changes were consistent across both coronal and sagittal brain sections. Notably, T cells exerted a pro-aging influence on nearby cells, often propagating their effects across longer spatial ranges than NSCs, which showed localized pro-rejuvenating effects.

In addition to cellular composition changes, gene expression also varied with age. For instance, microglia showed the largest number of age-related gene changes, particularly in immune response pathways. The study also identified specific patterns of gene expression changes across different brain regions, with white matter tracts exhibiting the largest changes. The findings emphasize immune-related genes increasing with age in microglia, contrasting with metabolic and developmental genes, which showed age-related declines.

To further explore the dynamics of aging, the researchers developed "spatial aging clocks" to predict the biological age of individual cells based on gene expression. This method accurately predicted cell age across various brain regions and cell types, including rare ones like NSCs and T cells. The clocks generalized effectively across sexes, datasets, and even other single-cell technologies, underlining their robustness.

The effects of rejuvenation interventions were also studied using the spatial aging clocks. Voluntary exercise and partial reprogramming were tested for their impact on brain aging. Exercise showed strong rejuvenating effects, particularly on brain vasculature, whereas partial reprogramming had more modest effects, notably rejuvenating NSCs and neuroblasts. Exercise had a broader impact, rejuvenating multiple cell types across brain regions, while partial reprogramming primarily benefited NSCs and neuroblasts with limited region-specific effects. Finally, the study examined how specific cells influence the aging of nearby cells, finding that T cells have a pro-aging effect, while NSCs have a pro-rejuvenating impact on neighboring cells.

Conclusions

This study offers high-resolution spatiotemporal profiling of the aging mouse brain, tracking gene expression across regions and cell types. By generating spatial aging clocks, it quantifies the effects of rejuvenating interventions and disease models. These clocks enable rapid assessment of aging and temporal processes at single-cell resolution. Importantly, the study demonstrates that T cells and NSCs play critical roles in modulating the aging process, influencing their neighbors through long- and short-range effects. The machine learning framework can be adapted to other tissues and species. The study also explores cell proximity effects, identifying potential mediators.