Everything Everywhere All at Once: Unraveling the Waves of Aging

Abstract

In a recent work reported in Science, Zhang et al. untangle dynamic changes arising across aging in multiple cell populations within thirteen organs using single-cell transcriptomics, and identify four distinct dynamic waves in which immune cells are the most affected populations.

Aging is considered the common factor behind many of the most prevalent diseases today, including cancer, neurodegenerative disorders or cardiovascular disease. Numerous efforts are underway to uncover common and targetable mechanisms underlying age-related diseases, with the goal of developing strategies to reduce their incidence and improve health outcomes in the later stages of life. However, the aging process is orchestrated by the interaction of various cell types across different organs simultaneously, making it highly complex to characterize the patterns that underlie the dynamics of this process.

In the 1990s, groundbreaking research in C. elegans unveiled that aging is malleable and dynamic with the identification mutations in genes such as Age-1 or Daf-2 that slowed down the aging process resulting in a two-fold life span increase. Genetic or pharmacological approaches in both Drosophila and mice can accelerate or delay the aging process. This body of research defined an initial set of nine hallmarks of aging, which provided a foundational framework for understanding the aging process. A decade of scientific advances in the field of aging has revised and expanded these hallmarks¹. In the last five years, the development and application of single-cell resolution transcriptome analysis has revolutionized the field of aging, enabling deeper understanding of the changes in cellular states and the different populations that emerge and disappear throughout the aging process².

Zhang et al³ used single-cell RNA sequencing technology to create an atlas of the changes occurring during aging in the cellular populations of 13 different organs - lung, heart, liver, kidney, muscle, stomach, duodenum, jejunum, ileum, colon, brown adipose tissue, inguinal adipose tissue and perigonadal adipose tissue - in both male and female mice. This “everything-everywhere-all-at-once” approach, called PanSci, resulted in the elaboration of a complex and dynamic landscape of the aging process with a methodology that bypasses the undesirable batch effect and increases the number of cells analyzed in other aging atlas, reaching a broader cellular diversity. The authors observed sex-specific differences in aging, with aged male livers upregulating genes related to acute-phase proteins, while female hepatocytes, more dependent on cholesterol metabolism, show increased expression of genes linked to fatty acid metabolism. Interestingly, the analyses revealed that among all cell types analyzed, immune cells – especially lymphocytes – exhibit the most changes in gene expression throughout aging.

Analysis of changes in population abundance across each organ identified four distinct waves during aging (Figure 1). The first one, occurring between 3 to 6 months of age, is characterized by the decline in the abundance of certain cell types such as Gmpr⁺ activated brown adipocytes or Dkk⁺ tenocytes in muscle. The second wave, taking place from 6 to 12 months of age, is marked by a drop in the numbers of CD4⁺ naïve T cells. The next stages – from 12 to 23 months – were defined by the expansion of immune cells, especially lymphocytes. Interestingly, the non-linear dynamic pattern seen in the PanSci analysis³ is observed also in the proteome of human blood samples from participants aged between 18 and 95 years, showing that the aging process occurs in a stepwise manner: first at the age of 34 years, then at 60, and finally at the age of 78⁴. These findings suggest that aging is not a linear process but rather occurs at varying rates, with certain ages undergoing more profound and significant changes than others.

Figure 1. The waves of aging.

Transcriptomic analysis of multiple tissues and different ages reveals four different waves in the aging process highlighting important changes in the immune cell compartment. Between 3 and 6 months there is a decrease of Gmpr⁺ adipocytes in brown adipose tissue and Dkk2⁺ tenocytes in muscles. In the second wave, occurring between 6 and 12 months, there is a marked decrease of naïve CD4⁺ T cells across various organs. The third wave, between 12 and 16 months, is characterized by increase of several immune cells such as age-associated B cells (ABC) expressing Tbx21 in lung and adipose tissues, CD8⁺ T cells expressing GzmK or Ly6c2 and Pde2a across different organs, or macrophages expressing Colq in the intestine. The last wave, after 16 months, is characterized by the increase of other immune populations such as Cd6⁺ Ccr9⁺ CD4⁺ T cells in the intestine or Pstpip expressing ABCs in lung and adipose tissues.

In the PanSci analysis³, the authors observed that immune cells, especially T cells and B cells, experience the most differential gene expression changes during aging, suggesting that lymphocytes are the most altered populations over time. These changes can be attributed to age-related physiological shifts affecting the immune system such as thymus involution, myeloid skewing of hematopoietic progenitors or the rise of terminally differentiated immune populations resulting from either homeostatic proliferation or clonal expansion upon antigen recognition.

The authors observed a decline in naïve T cells that begins in the second wave of aging, at 6-12 month of age. This is in accordance with observations made when examining the impact of different dietary restriction interventions to extend lifespan in a cohort of 960 genetically diverse mice, where the amount of naïve lymphocytes, both CD4⁺ and CD8⁺ T cells, positively correlated with longer life span⁵, highlighting that a high proportion of naïve T cell subsets is a top predictor of lifespan.

The third wave of aging is characterized by an increase in Cd8⁺ GzmK⁺ T cells in kidney, lung and adipose tissue. CD8⁺ GzmK⁺ T cells co-expressing exhaustion markers such as PD1 and Tox, and regulated by the transcription factor EOMES, promote paracrine senescence and systemic inflammation during aging⁶. GzmK secreted by this population of T cells cleaves some complement components, promoting the complement cascade activation and contributing to inflammatory diseases⁷.

Other populations augmenting during the third and fourth waves of aging are Cd4⁺ Cd6⁺ Ccr9⁺ and macrophages expressing Colq, both specific to the intestine; Cd8⁺ Ly6c2⁺ Pde2a⁺, that may result from an inflammatory environment or an aberrant activation of the CD8⁺ T cells; and two subsets of age-associated B cells (ABC), the first one characterized by the expression Tbx21, and the second one expressing Pstpip2. The ABCs are directly associated with humoral autoimmunity and promotion of inflammaging by producing inflammatory molecules such as interleukin (IL)-6 or interferon (IFN)-γ⁸. ABCs tend to switch their isotype toward IgG2a/c (IgG1 in humans)⁸, an immunoglobulin that is elevated in aged tissues. This immunoglobulin accumulates in sites with high abundance of senescence cells and may contribute to tissue aging⁹.

All these pieces of evidence support that the accumulation of age-associated immune cells, especially lymphocytes, contributes to inflammaging and tissue damage¹⁰. Genetic deletion of Ripk1, encoding a regulator of cell death pathways, in T lymphocytes induces senescence in CD4⁺ and CD8⁺ T cells resulting in premature aging and reduced lifespan¹¹. However, in PanSci analysis of 16-month-old lymphocyte-deficient mice (Rag1^-/- mice and Prkdc^-/- mice), the most affected age-associated subpopulations were lymphocytes, followed by Slco1a5⁺ kidney urothelial cells and Colq⁺ intestinal macrophages, suggesting a limited impact of lymphocytes on other non-lymphoid age-associated populations³. The authors suggest that this may be due to developmental defects in the mouse lacking functional lymphocytes from a very early developmental stage, but it may also be due to the complete absence of lymphocytes instead of the specific depletion of age-associated subpopulations, given the importance of some immune populations in aging such as naïve T cells⁵. Specific insight on the impact of lymphocytes in surrounding cells will require targeted depletion of lymphocyte subclusters and single-cell spatial transcriptomics in each organ. In a brain atlas spanning 20 different ages generated using single-cell spatial transcriptomics, the most altered genes during aging were associated with immune cells¹², with CD8⁺ cytotoxic T cells (and specifically IFN-γ) exhibiting the strongest pro-aging average proximity effect on their neighbors while the neural stem cells had the strongest anti-aging average proximity effect.

Collectively, Zhang et al. present a comprehensive, multi-organ transcriptomic atlas at single-cell resolution across five distinct age ranges in male and female mice, that identifies four waves of cellular abundance changes throughout aging and emphasizes the gene expression and abundance alterations of immune cell compartments. Their findings highlight the importance of understanding the impact of immune cells at non-immune organ sites and will present a key resource for advancing our understanding of immune function in diverse microenvironments during aging.