Unveiling the immune system aging in single-cell resolution

Abstract

This commentary investigates the findings presented in the article by Yang et al. in 2023, published in the Journal of Leukocyte Biology. This commentary first summarizes the spatial-temporal dynamics of regulatory T cells derived from mice (Tabula Muris Senis) of different ages (3, 18, and 24 mo) at different anatomic niches like lymph nodes and bone marrow. We also reported possible combinations of receptor–ligand interactions among T follicular regulatory cells, T follicular helper cells, and germinal center B cells, such as the calmodulin/Fas axis and PSGL-1/L-selectin axis. Then, we have elaborated on the significance of understanding aging regulatory T cells and offered some possible future research directions for Yang et al., contributing to a critical analysis of their recent study. Building on these foundations, further investigations and studies can be conducted to delve deeper into the mechanisms by which regulatory T cells influence health upon aging, potentially unveiling novel therapeutic targets to ameliorate age-related pathogenicity.

1. Introduction

Regulatory T cells (T-regs) are pivotal players in humoral immunity that serve as a “brake” to establish immunologic homeostasis. This can be achieved by generating cytokines such as TGF-β, IL-35, and IL-10, which inhibit effector T cells (ETCs).¹ Additionally, T-regs can suppress humoral immunity by triggering ETCs’ apoptosis and inducing dendritic cells to produce indoleamine 2,3-dioxygenase, which further suppresses ETCs.² Another proposed mechanism involves the generation of adenosine nucleosides, catalyzed by both CD39 and CD73.² T-reg–mediated immunosenescence is believed to be dependent on aging. Palatella et al.³ have recapitulated several detrimental effects of aging on T-regs. For instance, immunity against non-self-cells is generally weakened in the elderly.³ Although the mechanisms are intricate, the advancement of single-cell RNA sequencing (scRNA-seq) has revealed the transcriptional heterogeneity of T-regs, contributing to our understanding of aging immunity. Yang et al.⁴ investigated a relevant topic and published their findings in the current issue of the Journal of Leukocyte Biology.

In this commentary, we aim to summarize some main points covered by Yang et al.⁴ and provide comments. Furthermore, we propose future research directions for studying aging T-regs.

2. General introduction to scRNA-seq analysis

scRNA-seq analysis by the Seurat R package was employed by Yang et al.⁴ to study differentially expressed genes (DEGs) in mouse (Tabula Muris Senis) T-regs.⁵ The raw data were first obtained from the “Tabula Muris consortium,” a transcriptomic atlas for studying tissue aging in mice.⁶ Dead and doublet cells were filtered based on transcript abundance. The data set was normalized using the SCTtransform function to eliminate the effect of sequencing depth on T-reg heterogeneity.^4,7 Principal component analysis was conducted for dimensionality reduction of the data set. Then, cell clusters were identified using FindClusters and FindNeighbours functions, which were then projected on Uniform Manifold Approximation and Projection.⁴ DEGs in each cluster were identified by FindAllMarkers and quantified based on log₂(fold change) of transcripts.⁴

While scRNA-seq is powerful for understanding T-reg heterogeneity, it is still complicated by its technical limitations.⁸ First, scRNA-seq involves dissociation of viable cells by either enzymatic or heat treatment. Both require rigorous optimization to maintain cell viability.⁸ Such dissociation methods can be problematic as they potentially introduce unwanted transcriptional changes, causing errors in downstream analysis. Besides, scRNA-seq can introduce greater noise than bulk RNA sequencing, as each cell is independently considered a biological replicate.⁸

3. The main points covered by Yang et al.⁴

3.1 Spatial-temporal dynamics of age-dependent T-regs

Six clusters (C1–C6) of mouse-derived, organ-specific CD4⁺ T-regs have been identified through scRNA-seq as described.⁴ Most come from thymus, as indicated by the expression of thymic origin markers Ikzf2 and Nrp1.⁴

Cluster C1 can be further categorized into C1a and C1b, representing marrow-resident and 3 subgroups of lymphoid-resident T-regs, respectively.⁴ C1a has been found to upregulate expression of Sell, Pi16, Tgfbi, and Cpt1a, while C1b shows upregulation of Icos, suggesting that C1a cells are less mature than those in C1b.⁴ Generally, Prkch and Il27ra expressions are absent in C1. In C2, there are likely T-reg cells with compromised oxidative phosphorylation. C3 consists of CD4⁺ CD8⁺ thymocytes, which may serve as precursors of T-regs, as evidenced by the upregulation of Cd8a, Cd8b1, Dntt, and Rag1.⁴ T-reg cells in C4 and C5 have been identified as T follicular regulatory cells (Tfr), characterized by upregulated expression of Tfr marker genes Bcl6 and Cxcr5.⁴ T-reg proliferation in C4 is likely promoted by the upregulation of glycolytic genes.⁴ T-regs in C5 highly express CD150, suggesting a hematopoietic role in the marrow. Last, C6 has been identified as adipocyte-resident T-regs, as evidenced by the upregulation of Il10, Il1rl1, Pparg, and Gata3.⁴ Yang et al.⁴ also studied the temporal dynamics of C1–C6 using the propeller R package. The results indicated that the number of cells in C1 and C6 increases from 3 to 24 mo in mice, while the number of cells in C3 and C5 significantly decreases from 3 to 24 mo.⁴ Additionally, the number of cells in C2 and C4 decreases from 3 to 18 M and then increases from 18 to 24 mo.⁴

3.2 Interactomes analysis reveals possible therapeutic targets for the aging immune system

Given the immune-suppressive role of Tfr on germinal center (GC)–derived B cells and T follicular helper cells (Tfh), Yang et al.⁴ used the “iTALK” R package to investigate whether any age-dependent changes in combinations of receptor–ligand interactions occur between these cells.⁴ Interactome analysis revealed that PD-L1 on GC B cells of aging mice (18 mo) can suppress the development of PD-1–expressing Tfh cells and their homing to follicles.⁴ The interactome analysis also displayed possible mechanisms of Tfr-mediated regulation of Tfh apoptosis and exhaustion through the calmodulin/Fas axis and PSGL-1/L-selectin axis, respectively.⁴ Furthermore, possible pathways for Tfh activation during aging were discovered. In 18-mo-old mice, lymphotoxin alpha, which is essential for GC formation, was upregulated in Tfh cells.⁴Jag1 and Jag2, both of which encode ligands for Notch1 receptors, were upregulated in 24-mo-old mice.⁴ In addition, significant upregulation of Notch1 was identified in Tfh cells, indicating ongoing Tfh differentiation mediated by Notch signaling during aging.⁴

4. Significance of the findings

The aforementioned transcriptomic analysis reveals changes in T-reg subtypes during aging, providing valuable information for diagnostic markers in pathologic conditions. For example, T-reg accumulation in adipose tissue is correlated with insulin resistance and an increased risk of type 2 diabetes.^4,9 In addition, Safari et al.¹⁰ reported that STUB1 expression in T-regs inhibits the FOXP3-mediated anti-inflammatory program, potentially contributing to the development of rheumatoid arthritis. By studying transcriptional changes or gain/loss of T-reg subsets, we can gather more insights for diagnosing diseases and initiating early treatment.

The interactome analysis identifies receptor–ligand interactions among GC B cells, Tfr, and Tfh during aging.⁴ Such understanding enables the identification of potential drug targets, particularly proteins involved in cell signaling associated with aging T-regs.⁴ Researchers can then design therapeutic strategies, such as monoclonal antibodies or small molecule inhibitors, to disrupt these interactions. This approach holds the potential to reinvigorate aging immunity and improve immune function.

5. Comments on the work by Yang et al.⁴ and future research directions

Yang et al.⁴ have attempted to identify DEGs in clusters of T-regs derived from different lymphoid and nonlymphoid organs through scRNA-seq. However, such a tremendous load of scRNA-seq data may be a little bit overwhelming or even challenging for readers to understand the importance of each piece of data. What are the alternative hypotheses explaining how aging modifies the transcriptional signature of T-regs (Fig. 1)? What are the central questions to aging T-regs? What knowledge gaps in aging immunity do you want to fill? If these questions are not in mind beforehand, it is like finding a needle in a haystack. Perhaps we can narrow down the scope of the study if we can try to design experiments to test only a limited number of hypotheses.

As mentioned, the work by Yang et al.⁴ has recapitulated many gene signatures of aging T-regs. What is their relative biological significance in contributing to aging immunity? How can experiments be designed to validate the relative significance of certain gene sets in regulating aging immunity? We think gene perturbation assays could be useful for validating the importance of marker genes in driving aging immunity (Fig. 1). For instance, the original authors suggested that the calmodulin/Fas axis and PSGL-1/L-selectin axis are involved in promoting Tfh apoptosis and exhaustion, respectively.⁴ Indeed, knocking out (KO) these genes by the Cre-LoxP system in mice, followed by RNA sequencing of Tfr and Tfh from KO mice of different biological ages, can be conducted. This is useful for studying whether the loss of function of these genes can reverse the immunosuppressive role of Tfr on Tfh, potentially pinpointing targets for rejuvenating T-reg–mediated immunity upon aging. For instance, exhaustion markers (e.g. PD1, LAG3, TIGIT) in mice-derived Tfh can be studied to check if they are downregulated after KO immunosenescence-related genes.

Finally, Yang et al.⁴ mainly attempted to characterize the transcriptional signature of aging T-regs derived from mice. Despite the conservation of gene expression between mice and humans, it is better to study the gene expression of peripheral blood mononuclear cell (PBMC)–derived T-regs collected from donors of different biological ages (Fig. 1). Due to discrepancies between human and mouse genomes, resolving their transcriptomic differences provides a more solid foundation for further characterization of aging T-regs in clinical settings. Stratifying cohorts of PBMC donors is necessary such that the transcriptomic heterogeneity of T-regs is minimally confounded by factors such as sex, ethnicity, and health conditions of donors.

References

Author notes