Age-associated remodeling of T cell immunity and metabolism

Abstract

Aging results in remodeling of T cell immunity and is associated with poor clinical outcome in age-related diseases such as cancer. Amongst the hallmarks of aging, changes in host and cellular metabolism critically affect the development, maintenance, and function of T cells. Although metabolic perturbations impact anti-tumor T cell responses, the link between age-associated metabolic dysfunction and anti-tumor immunity remains unclear. In this review, we summarize recent advances in our understanding of aged T cell metabolism, with a focus on the bioenergetic and immunologic features of T cell subsets unique to the aging process. We also survey insights into mechanisms of metabolic T cell dysfunction in aging and discuss the impacts of aging on the efficacy of cancer immunotherapy. As the average life expectancy continues to increase, understanding the interplay between age-related metabolic reprogramming and maladaptive T cell immunity, will be instrumental for development of therapeutic strategies for older patients.

I. Introduction

The demographics of the human population is aging in an unprecedented manner. It is estimated that by the middle of the 21^st century, nearly one-quarter of the world population will consist of adults 65 years of age or older.¹ In addition to being associated with the development of multiple comorbidities, including diabetes, autoimmune disease, neurological deficits, and an increased susceptibility to infectious agents, aging is the most well-known risk factor for the development of cancer, the incidence of which disproportionately increases with age. It is now appreciated that functional dysregulation of the immune system is a central feature of the aging process. Surprisingly, our knowledge of the aging immune system, the mechanistic drivers of immune dysfunction during aging and the impact of the aged immune system on age-associated diseases remains incomplete.

The vertebrate immune system is equipped with a diverse cellular arsenal, capable of providing highly specialized and durable protection against various pathogens. Innate immunity provides immediate protection against invading pathogens and relies on a limited but evolutionary conserved set of germline-encoded receptors that recognize common constituents of microbes. Despite being rapid in nature, innate immune responses are limited in specificity, diversity and memory. By contrast, adaptive immune responses are highly specialized and rely on a diverse pool of lymphocytes capable of recognizing and responding to a large number of microbial and other antigens. The generation of a diverse adaptive immune repertoire is enabled by the rearrangement of antigen receptor genes resulting in different T and B lymphocyte clones with unique receptor specificities.

T lymphocytes originate from hematopoietic stem cell (HSC)-derived lymphoid progenitor cells in the bone marrow (Figure 1). These lymphoid progenitors migrate to the thymus, a specialized primary lymphoid organ critically involved in T cell development, and further mature into cytotoxic or helper T cell subsets. Upon thymic egress, mature T cells migrate into secondary lymphoid organs, where they continuously survey peptide fragments presented by major histocompatibility complex (MHC) molecules on the surface of antigen-presenting cells (APCs). Naïve T cell activation in the lymph node requires two signals: Ligation of the T cell receptor (TCR) by cognate peptide-MHC initiates naive T cell activation (signal one), and results in a downstream signaling cascade that promotes T cell proliferation, differentiation, and effector functions. Simultaneous engagement of the costimulatory receptor CD28 by its ligands B7–1 or -2 (CD80 or CD86) (signal two) is required for full T cell activation initiated by the TCR. In addition to these two signals, signal three provided by inflammatory cytokines including IL-12 and type I IFN promotes T cell differentiation and acquisition of effector functions. These signals culminate in the differentiation of activated T cells into specialized killer or helper T cell subsets (Figure 1).^2–4 While CD8⁺ T cells produce effector molecules including perforins and granzymes to directly induce destruction of target cells, subsets of CD4⁺ T cells play an important role in fine-tuning immune responses through the secretion of specialized cytokines and by engaging in interactions with various immune cells.

Figure 1: A brief overview of T cell development and activation.

T cell development begins in the bone marrow with hematopoietic stem cells that acquire a lymphoid progenitor state. These cells migrate to the thymus, a specialized primary lymphoid organ, to undergo additional maturation and selection against self-reactive T cell clones. During this process, progenitor T cells differentiate into either mature CD4⁺ or CD8⁺ T cells and exit the thymus. These antigen-naive T cells can be activated in the lymph node by dendritic cells which provide TCR, co-stimulatory and cytokine signaling for the acquisition of effector T cell functions.

Although much of our understanding of T cell immunity originated from studies utilizing immunocompetent young hosts, studies in aged model systems have taught us that T cell function and phenotype are considerably altered with age.⁵ The complex biological changes associated with aging impact multiple facets of T cell immunobiology including development, activation, differentiation, and survival. Age-associated deficits in T cell responses against cancer, infectious and autoimmune diseases are well documented in elderly demographics. Recent studies have revealed that T cells a well as B cells play an important role in shaping the immune response to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections.⁶ Furthermore, one of the most critical determinants of disease severity was age,⁷ and studies have highlighted the crucial role of age-associated T cell dysregulation in the perpetuation and aggravation of disease severity.⁸ However, the mechanisms driving age-associated T cell dysfunction and disease progression remain an active area of investigation.

Impaired T cell function in aged hosts is attributed to the cumulative deficits in cellular and physiological integrity caused by molecular hallmarks of aging. Genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, cellular senescence and numerous other biological deficits all contribute to cellular and physiological decline.⁹ Amongst these perturbations, one of the most prominent hallmarks of aging involves metabolic dysfunction in which chronic dysregulation of cellular and host bioenergetic programs results in altered cellular function and increased susceptibility to metabolic disorders, including obesity, type II diabetes and non-alcoholic fatty liver disease.¹⁰ These metabolic adaptations may be driven by age-associated defects in mitochondria, which are multi-faceted organelles that play a critical and non-redundant role in the production of ATP, maintenance of calcium homeostasis and regulation of cellular apoptosis.¹¹ Age-associated mitochondrial dysfunction results in impaired mitochondrial respiration and mitophagy, augmentation of reactive oxygen species (ROS) production, and increased apoptosis in inflammatory settings. The relationships between mitochondrial degeneration and aging have been the subject of many recent investigations.¹²

Age-associated mitochondrial and metabolic reprogramming has a direct impact on the immune system and its role in diseases such as cancer, autoimmunity and chronic infections. In particular, the field of T cell metabolism has become an area of substantial investigation over the past decade. Recent work highlights the importance of cellular bioenergetic programs in peripheral T cell maintenance, activation, differentiation, and exhaustion. However, the metabolic pathways which control the fate of T cells are often rewired as a result of age-associated mitochondrial and metabolic perturbations.¹³ In this review, we summarize past progress and recent advances in our understanding of the role of aging in T cell immunity and metabolism, with emphasis on the implications of metabolic reprogramming of aged T cells in regulating anti-tumor immunity.

II. Early studies of T cell aging

The notion that aging impairs T cell development and function is a long-standing observation. Studies in the 1960’s revealed that aging negatively impacts the activation and distribution of T cells. In humans, both the frequency and absolute quantity of spontaneous sheep red blood-cell (SRBC) rosette-forming T lymphocytes decreased with age,¹⁴ and was correlated with an overall decline in thymosin, a 5-Da polypeptide hormone critically involved in T cell development and function.¹⁵ These findings were consistent with impairment of delayed hypersensitivity responses (DHR) in aged individuals upon antigenic sensitization with 2,4-dinitrochlorobenzene (DNCB).^16,17 Furthermore, in vitro stimulation of lymphocytes using mitogens such as phytohemagglutinin (PHA) resulted in attenuated T cell proliferative responses in aged individuals.^18–20 Similar numerical and functional T cell deficits have been documented in aged mice.^21–24 Studies have highlighted the potential implications of age-associated T cell dysfunction in various disease settings,^25–27 underscoring a need to understand the cellular and molecular mechanisms involved.

Building upon the previously acknowledged impact of aging on T cell immunity, recent studies have integrated more sophisticated tools such as single-cell DNA/RNA sequencing to further profile the systemic effects of aging on the immune system.^28,29 These efforts, along with many other studies, have provided greater insight into how aging modulates T cells. It is now appreciated that numerous age-associated biological alterations may contribute to a defective T cell response.⁵ However, the biological mechanisms underlying age-associated T cell dysfunction remain unclear and context dependent.

III. T cell development

One of the most striking features of aging is thymic involution, during which the thymus degenerates over time and no longer replenishes naïve CD8⁺ and CD4⁺ T cell pools necessary for protective immunity.^30–32 Impaired naïve T cell development disrupts the composition of peripheral T cells, resulting in compensatory expansion of pre-existing T cell clones. These adaptations lead to decreased TCR diversity, an altered distribution of T cell subsets, and qualitative changes in T cell responses following TCR stimulation. These cellular changes and clonal expansion may contribute to the loss of immune responses to a diverse set of antigens in the aged.

In early studies, thymectomy of young C57BL/6 and A/J mice accelerated the decline in cellular immunity and precipitated the generation of dysfunctional T cell populations characterized by attenuated responsiveness to PHA and staphylococcal enterotoxin B (SEB), similar to those observed in aged mice.²⁴ Consistent with this finding, altered lymphocyte sub-type distribution and impaired responsiveness to mitogenic signaling is observed in patients following thymectomy.^24,33 Species-specific adaptation to thymic involution also has been reported. In contrast to mice, which predominantly rely on thymically-derived T cells to replenish the peripheral naïve T cell pool, peripheral expansion of naïve T cells in aged humans was sufficient to compensate for thymic involution.^34–36 This divergence suggests that mouse studies may offer limited insight into the processes governing adaptation to thymic involution in humans and the overall impact of such adaptation on the human aging process. While the use of aged mice as an experimental model system will continue to guide our basic understanding of immune aging, the integration of human studies, including use of primary human cells and grafts, will be necessary for the validation of mechanistic insights obtained from animal models.

Although several mechanisms have been implicated in age-associated thymic involution, the central driver of thymic attrition remains unclear. The process of aging modifies hematopoietic reprogramming of precursor cells in the bone marrow to favor commitment towards myeloid over lymphoid lineages³⁷ which may contribute to reduced thymic output. However, thymic atrophy is initiated prior to alterations in the hematopoietic compartment. Studies have demonstrated that age-associated thymic atrophy in healthy hosts is associated with alterations in thymic stromal cells^38,39 and that attenuated H₂O₂-reducing enzyme and mitochondrial dysfunction of stromal cells may contribute to thymic atrophy.⁴⁰ Consistent with this finding, genetic and biochemical augmentation of anti-oxidants was shown to ameliorate thymic atrophy.⁴⁰

The metabolic environment plays a major role in thymic atrophy and generation of naïve T cells. For example, age-associated thymic involution correlates with accretion of lipid-laden adipocytes, while disrupting accumulation of thymic adipocytes through caloric restriction attenuates tissue atrophy.⁴¹ A recent study by Spadaro et al. demonstrated that caloric restriction in humans is associated with impaired expression of platelet activating factor acetyl hydrolase (PLA2G7), increased mitochondrial biogenesis and PPAR-α-driven lipid oxidation in thymic adipocytes. Deletion of PLA2G7 in mice resulted in decreased thymic lipoatrophy.⁴² Consistent with the deleterious effects of age-associated adipocyte remodeling of thymic tissue architecture, high-fat diet (HFD)-induced obesity accelerated thymic involution, reduced peripheral naïve T cell numbers and decreased TCR repertoire diversity in mice,⁴³ while progressive adiposity in mid-aged humans reduced thymic output.⁴³ Similarly, severe thymic atrophy was observed in leptin-deficient obese mice in which reconstitution of leptin reversed obesity and thymic involution.^44,45 In contrast to these findings, other work suggests that thymic size, cell number and output may increase in HFD-induced obese mice,⁴⁶ as a result of the role of leptin in augmenting thymocytes and T cell function.^47,48 Additional studies are needed to better understand mechanisms connecting nutrient signaling with T cell development during aging. In summary, age-associated metabolic alterations significantly impact thymic structural integrity and T cell development, which further impact T cell responses required for protective immunity.

IV. Naïve T cell homeostasis

In the periphery of aged hosts, pre-existing T cells undergo additional expansion to compensate for impaired naïve T cell development. During this process, aging exerts distinct effects across pre-existing naïve CD8⁺ and CD4⁺ T cells. For reasons incompletely understood, naïve CD8⁺ T cells exhibit greater vulnerabilities to the effects of aging in comparison to CD4⁺ T cells.^49–53 Consistent with this finding, one of the cardinal features of aging in healthy adults is reduced naïve CD8⁺ T cell numbers in the periphery.⁵⁴ The presence of age-associated “memory T cells with a naïve phenotype” (T_MNP cells) in the naïve CD8⁺ T cell pool suggests that the number of naïve CD8⁺ T cells is likely overestimated in most studies which do not utilize T_MNP markers such as CD49b and CXCR3 for additional segregation.⁵⁵ Overall, the aging process preferentially decreases the number and longevity of CD8⁺ T cells despite increasing the longevity of naïve CD4⁺ T cells.⁵⁶ These differences may be explained by the presence of naïve-like CD45RA⁺CD25^dimCD4⁺ T cells in aged mice, which are characterized by increased sensitivity to IL-2.⁵¹ Although the consequences of aging on naïve CD4⁺ T cell function remain unclear in vivo, several studies have demonstrated defective proliferation and impaired CD25 expression in aged naïve CD4⁺ T cells following TCR engagement in vitro.^56,57

Defects in the naïve CD8⁺ T cell pool can be attributed to cell-extrinsic and -intrinsic changes that accompany the aging process. Cell-intrinsic factors, including genetic, epigenetic, and metabolic alterations, cumulatively contribute to the overall dysfunction of naïve CD8⁺ T cells. Telomere length is significantly reduced in the naïve CD8⁺ T cell compartment, likely as consequence of heightened homeostatic proliferation in aged hosts.^58,59 Telomere shortening of CD8⁺ T cells is augmented with age-associated elevation of mitochondrial stress and ROS accumulation. Mitochondrial dysfunction-induced telomere damage was reversed with the use of a ROS scavenger.⁶⁰ Decreased expression of the double-strand break repaired nuclease MRE11A causes telomeric damage and accelerates the aging process in T cells from patients with rheumatoid arthritis.⁶¹ Telomere shortening disrupts T cell homeostasis and survival, as illustrated by the finding that telomerase mutation carriers with short telomere length develop T cell immunodeficiency.⁶² Conversely, increased telomere length and telomerase activity in T cells has been associated with health benefits in healthy centenarians.⁶³

Age-associated epigenetic modifications also contribute to remodeling of the naïve CD8⁺ T cell compartment. Aged naïve CD8⁺ T cells exhibit differential chromatin accessibility profiles compared to aged naïve CD4⁺ T cells.^52,64 Transcriptional activity of NFκB, STAT, YY1, TCF1 and NRF1, all of which play key functions in T cell proliferation, were reduced in aged naïve CD8⁺ T cells.^52,64,65 For example, aged naïve CD8⁺ T cells exhibit decreased binding of NRF1 to its motif as a result of limited chromatin accessibility.⁶⁶ With the transcriptional coactivator PGC-1α, NRF1 promotes transcription of genes involved in glutathione biosynthesis,^67,68 the electron transport chain (ETC),⁶⁹ and positively regulate the mitochondrial transcription factor A (TFAM),⁷⁰ all of which play critical roles in mitochondrial biogenesis and respiration. Consistent with these data, basal and maximal oxygen consumption rates (OCR) were decreased in aged naïve CD8⁺ T cells despite observations of increased mitochondrial mass.⁶⁶ Naïve CD8⁺ T cells heavily rely on oxidative phosphorylation (OXPHOS) for homeostatic expansion, and age-associated epigenetic regulation of NRF1 and genes associated with mitochondrial respiration likely contribute to defective homeostatic proliferation of aged naïve CD8⁺ T cells. Lastly, although epigenetic changes directly impact T cell metabolism, metabolic changes also influence the epigenetic states of T cells, ultimately shaping T cell fate decisions and differentiation programs.^71–74 However, it remains unclear whether metabolic changes contribute to altered epigenetic states in aged T cells.

Post-transcriptional regulators, including microRNAs, also play an important role in aged-associated naïve T cell remodeling. In particular, age-dependent reconfiguration of microRNA networks predicted a major defect in FoxO1 signaling,⁷⁵ which has been reported to repress T cell senescence and promote proteostasis.^76,77 Furthermore, FoxO1 directly promotes expression of the interleukin 7 (IL-7) receptor,⁷⁵ which plays an important role in maintaining naïve CD8⁺ T cell homeostasis.⁷⁸ Consistent with this finding, IL-7R expression was reduced in naïve CD8⁺ T cells from aged hosts and inhibition of FoxO1 in young naïve CD8⁺ T cells decreased IL-7R expression to levels observed in aged T cells.⁷⁵ Although IL-7 modulates glucose uptake and glycolytic flux, reports on the role of IL-7 in the regulation of Glut1 expression are conflicting.^79,80 IL-7 also increases the expression of CD98, a heterodimer of SLC3A2 and SLC7A5 which forms the large neutral amino acid transporter, to sustain IL-7 induced T cell growth.⁸¹ While age-associated changes in metabolic, genetic, epigenetic, and post-transcriptional regulation have potential to contribute to defective expansion of naïve CD8⁺ T cells in aged hosts, there is still a need to investigate how these changes impact the function of aged T cells to alter immune control in vivo.

V. T cell activation & metabolism

Naïve T cell activation is a highly regulated process which relies on appropriate integration of TCR, co-stimulatory and cytokine signaling. Early studies suggested that naïve T cell activation becomes dysfunctional with age, as evidenced by decreased proliferation in response to T cell stimulation.^18–20 Although the consequences of aging on early events following TCR stimulation are poorly characterized for both CD8⁺ and CD4⁺ T cells, several distinct molecular and cellular features of aged T cell activation have been described. Activation of aged naïve T cells is associated with impaired immune synapse formation,⁸² recruitment of phosphorylated Lck and LAT to the lipid raft,⁸³ Ca²⁺ flux,^84,85 cytokine production, and expression of activation markers such as CD25 and CD69.⁸⁶ Moreover, TCR engagement of aged T cells is associated with increased apoptosis and DNA damage response.^87,88 As previously described, prior to TCR engagement, aged naïve CD8⁺ T cells display baseline alterations in genetic, epigenetic, and post-transcriptional regulatory networks which have important implications for T cell activation. For example, the microRNA mir-181a controls TCR activation threshold during thymic selection and peripheral T cell activation. Loss of miR-181a expression during aging results from the age-associated decline of YY1, a transcription factor that promotes miR-181a expression by controlling pri-miR-181a/b1 enhancer activity.⁸⁹ In addition to baseline alterations in TCR signaling, metabolic perturbations also contribute to defective cellular activation.

Metabolic requirements for T cell activation

Metabolic reprogramming following T cell activation is indispensable for maintenance of high energetic demands resulting from rapid proliferation, differentiation and acquisition of effector T cell function.⁹⁰ Naïve T cells exist in a state of quiescence characterized by low metabolic activity and preferential use of OXPHOS for the generation of ATP. Following TCR engagement, naïve T cells rapidly switch towards the use of aerobic glycolysis, with a reliance on glutaminolysis, pentose-phosphate pathway activity, and 1-carbon metabolism to increase energy production.^91–93 While numerous bioenergetic pathways are actively utilized during T cell activation, we focus here mainly on glycolytic and mitochondrial metabolism which are highly altered with aging (Figure 2).

Figure 2: Naïve T cells from aged hosts display attenuated glycolysis and mitochondrial metabolism.

Aged naïve T cells display defective TCR signaling (a), early metabolic activities (including glycolysis and pentose phosphate pathway) (b), and transcriptional and proteomic reprogramming required for mitochondrial biogenesis (c).

Early TCR stimulation events induce aerobic glycolysis, leading to rapid production of ATP and cytokine production.^94,95 Aerobic glycolysis involves the activation of pyruvate dehydrogenase kinase 1 (PDHK1) which phosphorylates and deactivates pyruvate dehydrogenase (PDH). Increased PDHK1 and subsequently decreased PDH activity in T cells lead to attenuated conversion of pyruvate into acetyl-coenzyme A for mitochondrial import and increased conversion of pyruvate into lactate, catalyzed by lactate dehydrogenase (LDH).⁹⁵ This metabolic switch disengages the binding of LDH to the 3’ untranslated region (UTR) of IFN-γ, IL-2 and TNF-α mRNA and thereby increases cytokine production. Lactate dehydrogenase A (LDHA) also controls IFN-γ expression through an alternative mechanism involving maintenance of acetyl-CoA, which enhances histone acetylation and transcription of IFN-γ mRNA.⁹⁶ Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) is another crucial enzyme upstream of the glycolytic pathway which binds to the 3’ UTR of IFN-γ mRNA prior to T cell activation, and TCR-mediated glycolysis disengages GAPDH-IFN-γ mRNA binding to promote IFN-γ translation.⁹³ In conjunction with regulating cytokine production, glycolysis plays a crucial role in the generation of metabolic fuels required for optimal T cell activation, including the biosynthesis of nucleotides, amino acids, proteins and fatty acids.⁹⁷ Glycolysis and glutaminolysis were also shown to co-operatively deprive the hexosamine pathway of glucose and glutamine, which inhibits N-glycan branching and favors the growth and differentiation of pro-inflammatory T_h17 cells.^98,99 Lastly, glycolysis increases levels of acetyl-CoA and promotes histone acetylation, thereby enabling the generation of open and permissive chromatic structures required for T cell activation.^73,100 Together, the glycolytic pathway initiates metabolic reprogramming to enable rapid cytokine production, proliferation, and activation of T cells.^97,101,102

Upon TCR-stimulation, mitochondria buffer Ca²⁺ flux in the cytosol.^103,104 Mitochondrial localization proximal to the immune synapse decreases local cytosolic Ca²⁺ and calcium-dependent channel inactivation, thus augmenting overall calcium influx.^105–107 Uptake of Ca²⁺ into the mitochondrial matrix enhances the activity of proteins associated with TCA cycle, and promotes OXPHOS, ATP generation and ROS production, all of which have an important role during T cell activation.^106,108 The initial Ca²⁺ flux followed by Ca²⁺ influx from the ER and extracellular compartment activate calcineurin, which leads to the activation of NFAT. Mitochondrial production of ROS plays an important role in inducing NFAT-dependent Myc expression, which augments T cell proliferation and metabolic reprogramming.^91,108 In particular, Myc-dependent metabolic reprogramming involves augmentation of glycolysis and glutaminolysis.⁹¹

Mitochondrial biogenesis and remodeling accompany TCR stimulation, and is accompanied by remodeling of the mitochondrial proteome and augmentation of 1-carbon metabolism.^92,109 Mitochondrial respiration and ATP production also play a critical role in supporting energy demands following T cell activation and inhibition of ATP synthase using oligomycin blunts T cell activation.⁹³

During mitochondrial biogenesis, mitochondria are rewired to become anabolic engines. One-carbon metabolism utilizes serine to generate one-carbon units for thymidine and purine biosynthesis, which play an important role in early T cell activation and proliferation (Figure 2).^92,110 During this process, serine is converted to glycine in the cytosol and mitochondrial matrix by serine hydroxymethyltransferase-1 and 2 (SHMT1 and 2), respectively. In addition, serine-driven one-carbon metabolism contributes to the maintenance of redox homeostasis through the generation of glycine for glutathione synthesis and NADPH, which supports antioxidant defence and reductive biosynthesis.^92,111 In summary, a metabolic switch towards aerobic glycolysis provides T cells with initial energy and cues required for rapid cellular proliferation, but mitochondrial metabolism is required for sustaining T cell activation.

Age-associated metabolic changes in T cell activation

Although the literature broadly suggests that age-associated metabolic dysfunction may contribute to an altered T cell response, it remains unclear precisely how age-associated metabolic alterations impact the bioenergetic program of T cells at discrete phases of development and activation. Here, we focus on age-associated metabolic alterations upon naïve T cell activation.

Alterations in glycolysis correlate with age-associated T cell dysfunction. Early studies suggested that LDH activity was significantly reduced in aged T cells and correlated with decreased T cell proliferation.^33,112 Consistent with these studies, activation of naïve aged CD4⁺ T cells resulted in attenuated ECAR and reduction in metabolites associated with glycolysis and the pentose phosphate pathway (Figure 2).⁵⁷ It remains unclear whether glycolysis is also impaired in naïve CD8⁺ T cells during aging. Reduced glycolysis in CD4⁺ T cells may result from age-associated changes in the TCR signaling threshold. Age-associated downregulation of mir-181a in CD4⁺ T cells may contribute to decreased TCR sensitivity.^65,89,113 Mir-181a is a negative regulator of dual specificity phosphatase 6 (DUSP6) and SIRT1, which in turn function as negative regulators of T cell activation.^65,89 Impaired mir-181a activity increases the inhibitory functions of DUSP6 and SIRT1, ultimately leading to decreased T cell activation. Furthermore, SIRT1, a NAD⁺-dependent protein deacetylase, targets various signaling molecules, histones and transcription factors which are important for metabolic functions. SIRT1 attenuates glycolysis through deacetylation of phosphoglycerate mutase-1 (PGAM-1)^114,115 which catalyzes the conversion of 3-phosphoglycerate into 2-phosphoglycerate to fuel glycolysis during T cell activation.⁹⁴ Further work is needed to better understand metabolic alterations and how they impact early aspects of aged T cell activation.

Mitochondrial dysfunction is one of the most prominent hallmarks of aging,⁵ and alters the bioenergetic program in T cells. Early studies revealed that mitochondrial morphology is drastically altered in aged T cells.^116–118 Consistent with these observations, activation of aged T cells resulted in smaller mitochondria characterized by decreased mitochondrial respiration and ATP production.^57,119 Mitochondrial dysfunction in aged T cells can be partially attributed to defective 1-carbon metabolism which impairs purine and thymidine biosynthesis necessary for optimal T cell proliferation and survival (Figure 2).⁵⁷ Nakajima et al. also demonstrated that aged naïve CD8⁺ T cells are less capable of generating CD44^lowCD62L^low cells upon T cell activation; CD44^lowCD62L^low cells represent a transitional state of early activation and require 1-carbon metabolism to acquire an effector T cell phenotype.¹²⁰ Activation of aged CD8⁺ T cells was associated with decreased 1-carbon metabolism-related genes including shmt2. Notably, activation of aged naïve T cells with products of 1-carbon metabolism including formate and glycine partially rescued the defects observed in aged T cells,⁵⁷ demonstrating the import of 1-carbon metabolism and indicating that other pathways are also involved with age-associated T cell dysfunction.

Alterative pathways may also contribute to age-associated reduced purine biosynthesis. Aging is associated with increased hypoxanthine-guanine phosphoribosyl-transfer (HGPRT) mutations in T cells;^88,121 HGPRT plays a crucial role in generating purine nucleotides via the purine salvage pathway. Impaired activation of aged T cells is also associated with decreased purine nucleoside phosphorylase (PNP) activity.¹²² PNP, which is upregulated upon TCR stimulation, plays a critical role in generating purine nucleotides through the salvage pathway, and additionally promotes T cell proliferation and effector function.¹²³ Thus, aging may attenuate purine biosynthesis by abrogating both 1-carbon metabolism and the nucleotide salvage pathway (Figure 2). Beyond defective 1-carbon metabolism and nucleotide biosynthesis, a recent study showed that age-associated accumulation of C14/C16 ceramide in mitochondria induces T cell mitophagy which subsequently leads to mitochondrial dysfunction and decreased ATP production.¹¹⁹ While aging leads to defective mitochondrial function during T cell activation and restrains effector T cell differentiation, further studies are needed to identify causal metabolic pathways that may be targeted to attenuate age-associated immune cell dysfunction.

VI. Age-associated changes in T cell sub-types

Aging is associated with unique T cell subsets that are otherwise absent or relatively scarce in healthy young hosts (Figure 3). These T cell subsets often display evidence of cellular senescence or exhaustion and contribute to dysfunctional immune responses. A better understanding of the metabolic mechanisms involved in the formation, maintenance and functional activity of these subsets may reveal novel targets for therapeutic interventions aimed at reversing T cell dysfunction in aged individuals.

Figure 3: Reconfiguration of T cell subtypes with aging.

A. Aging is associated reconfiguration of immune cell types in the periphery, including the expansion of myeloid cells, decreased proportions of naïve CD8⁺ T cells, and accumulation of unique age-related T cell types. B. Aged naïve T cells undergo defective catabolic and anabolic metabolism upon TCR-stimulation, resulting in decreased functional effector and memory T cells. C. Trajectory of naïve T cells in aged hosts. As a result of cell-intrinsic and environmental perturbations, naïve T cells differentiate and expand into age-related T cell populations: CD8⁺CD28⁻ T cells, terminally differentiated effector memory T cells re-expressing CD45RA (Temra), age-associated T (Taa) cells and virtual memory T (Tvm) cells have been reported to expand with aging and display distinct immune and metabolic signatures.

Following effector T cell differentiation, T cells adopt a distinct metabolic configuration as they become memory cells capable of providing long-term protective immunity.^90,124 Memory T (Tm) cells are characterized by reduced aerobic glycolysis and rely on fatty acid oxidation (FAO). In addition, Tm cells exhibit a marked increase in the numbers of fused/elongated mitochondria as well as a heightened spare respiratory capacity.^{101,125–127} Although these changes maintain Tm cells in a state poised for rapid response following re-encounter of antigen, they also serve as vulnerabilities and targets of dysregulation in the aged setting. As such, impaired proliferation, cytokine production and cognate helper function are all features of memory T cells originating from aged naïve CD4⁺ T cells (Figure 3).¹²⁸ In an adoptive transfer model, infusion of memory T cells from aged mice into young hosts did not rescue age-associated T cell dysfunction, suggesting that defects in aged memory T cells are intrinsically imprinted.¹²⁹

Despite the deficits in Tm cell differentiation and function, the proportion of antigen-experienced T cells is increased in aged hosts, including T cells expressing CD44 in mice,¹³⁰ and CD45RO in humans.¹³¹ CD44 expression is broadly utilized to distinguish antigen-experienced effector and memory T cells from naive T cells. Among the antigen-experienced pool, T cells with self-reactive TCRs exhibit preferential accumulation with aging, as a result of selective survival.^132,133 Aged CD44⁺ T cells often display traits associated with dysfunctional T cells. For example, CD44^hiCD8⁺ T cells from aged mice exhibit higher expression of co-inhibitory receptors including PD-1, LAG3, 2B4 and CD160, and display features of exhausted T cells.^134,135 T cell exhaustion arises in the setting of persistent antigen stimulation, such as chronic infections and cancer, and is characterized by progressive loss of effector function, expression of co-inhibitory receptors, a distinct epigenetic state, and arises in the settings of persistent antigen stimulation, such as chronic infections and cancer.^136,137 In the following sections, we summarize key metabolic and immunologic features of age-related T cell subsets, including attributes associated with cellular senescence.

Senescence and quiescence

It is important to note that the terms “senescence” and “quiescence” are often used interchangeably in the aging literature. Whereas quiescence represents a reversible state of cell cycle arrest, often resulting from the lack of growth factors and mitogenic signals, senescence represents a degenerative cell state characterized by stable proliferative arrest that is unresponsive to mitogenic stimulus.¹³⁸ Presently, the biology and distinct phenotypes of quiescent and senescent T cells in aging are not well defined and will be an important area of future investigation. Potential markers of T cell senescence include senescence-associated β-galactosidase (SA-β gal), senescence-associated secretory phenotype (SASP), CDK inhibitors p21/p16 and the DNA damage response histone H2A family H2AX (γH2AX), amongst others.^139,140 In human CD4⁺ T cells, inhibition of the respiratory chain complex 1 with low-dose rotenone accelerated the acquisition of a senescent state, suggesting that mitochondrial fitness governs the susceptibility towards immunosenescence.¹⁴¹

Recent studies highlighted a role for senescent immune cells as drivers of the aging process.^142–144 Deletion of Ercc1 in HSCs, which encodes a protein crucial for DNA damage repair, resulted in a premature onset of immunosenescence and multiorgan aging pursuant to attrition of the immune compartment.¹⁴² Mice deficient in perforin, a glycoprotein required for T cell killing, exhibit higher multi-organ senescent cell burden and suffer from age-associated morbidities, collectively suggesting that impaired immune surveillance is sufficient to contribute to aging and age-associated pathologies.¹⁴⁵ It is important to note that the above studies rely on genetic strategies to accelerate aging or induce senescence in immune cells and may not reflect the natural aging process. Thus, there is a need to identify features of T cell senescence and the contribution of these cells on systemic aging.

Senescent-like CD8⁺CD28⁻ T cells

CD8⁺CD28⁻ T cells, a heterogenous population of antigen-experienced memory and senescent T cells, accumulate with age (Figure 3).^146,147 At birth, the majority of CD8⁺ T cells express CD28 whereas the proportion of CD28-negative cells increases substantially by the age of 80.¹⁴⁶ These cells are primarily a feature of human and non-human-primates but not mice.¹⁴⁸ CD8⁺CD28⁻ T cells develop and accumulate with aging as a result of repeated antigen exposure^149,150 and cytokine signaling resulting from common γ-chain cytokines (IL-7 and IL-15), type-I interferons (IFN-α and IFN-β) and TNF-α.^151,152 CD8⁺CD28⁻ T cells exhibit evidence of cellular senescence including impaired TCR-mediated proliferation,¹⁵³ decreased telomerase activity^154,155 and augmented apoptosis.¹⁵¹ Nevertheless, CD8⁺CD28⁻ T cells are capable of rapid cytokine-mediated oligoclonal expansion^152,156 and produce high levels of cytokines such as TNF-α and IFN-γ.^157,158

CD8⁺CD28⁻ T cells display distinct metabolic properties compared to CD8⁺CD28⁺ T cells. CD28⁻ T cells express low levels of SIRT1, consistent with the association between downregulation of SIRT1 in aging and cellular senescence.^159,160 In metabolic organs such as white adipose, liver and skeletal muscle, SIRT1 promotes the activity of the nuclear receptor and transcriptional co-activator PPAR-γ and PGC-1α respectively, to upregulate mitochondrial biogenesis, and to mobilize glucose and fatty acids.^161,162 Despite the importance of SIRT1 in mitochondrial biogenesis and function, comparable levels of mitochondrial respiration were observed between CD8⁺CD28⁺ and CD8⁺CD28⁻ T cells. CD8⁺CD28⁻ T cells exhibited augmented basal glycolysis and a heightened glycolytic capacity as a consequence of the lack of SIRT1-mediated deacetylation of FoxO1, which is required to prevent proteasomal degradation of FoxO1.¹⁵⁹ The overall increase in glycolysis in CD8⁺CD28⁻ T cells likely contributes to cytokine-mediated oligoclonal expansion and increased cytokine production.

The functional consequence of CD8⁺CD28⁻ T cell accumulation in human disease appears to be context-dependent. Initial reports linked the accumulation of CD28-negative T cells with aggravation of disease outcome in the settings of cancer, infectious, and autoimmune disorders.¹⁶³ However, additional studies suggest CD8⁺CD28⁻ T cells may exhibit protective effects and tolerogenic features, with regulatory function being ascribed to this age-associated subset.^164,165 These paradoxical findings may be explained by the heterogeneity of the CD8⁺CD28⁻ pool, with either immune-stimulatory or suppressive populations dominating in a disease-specific context.

Terminally differentiated effector memory T (Temra) cells

CD45RA⁺CD27⁻ Temra cells in humans, characterized by expression of the naïve T cell marker CD45RA, are a unique cell type distinct from effector memory T cells,^166,167 and accumulate and undergo functional changes with age. These cells contain a high percentage of CD28-negative cells and likely represent a portion of the CD8⁺CD28⁻ cell population which accumulates with age.¹⁶⁸ Similar to CD8⁺CD28⁻ T cells, Temra cells exhibit features of cellular senescence including secretion of SASP, activation of p38 mitogen-activated protein kinase (MAPK), phosphorylation of γH2AX, and expression of surface markers associated with senescence such as KLRG1 and CD57.^169–171 It remains unclear whether Temra can be described as senescent memory cells, given that cellular senescence is associated with defective replication potential and Temra proliferation can be rescued through inhibition of p38 MAPK pathway or with IL-15 supplementation.^172–174

CD8⁺ Temra cells produce higher levels of TNF-α and IFN-γ and display decreased proliferation compared to other types of memory T cells (Figure 3).^167,175 These cells possess metabolic signatures of decreased mitochondria, OCR, and mitochondrial membrane potential, but increased mitochondrial and cytosolic ROS, and anaerobic glycolysis.^171,173 Mitochondrial dysfunction in Temra cells is mediated through activation of p38 MAPK, which inhibits mTOR-independent autophagy and subsequent production of metabolic precursors.^171,172 In CD8⁺ Temra cells, glucose and fatty acid uptake is defective and contributes to a reduced mitochondrial content, OXPHOS and ATP production.¹⁴¹ Temra cells are known to accumulate in specific chronic infections such as cytomegalovirus infection,^176,177 however the functional consequence of CD8⁺ Temra cell expansion in age-associated diseases remains unclear.

Virtual memory T (Tvm) cells

Tvm cells represent another T cell subclass that accumulates with age in mice (Figure 3).¹⁷⁸ These cells, characterized as CD44^hiCD49d^low, are distinct from other memory T cell subpopulations by virtue of their foreign antigen-naïve state.^179,180 Tvm cells originate from a self-reactive T cell population which expands peripherally in response to elevated cytokine signaling, including signals derived from type I interferons, IL-4 and IL-15. Accordingly, Tvm cells express high levels of receptors for homeostatic cytokines, including IL-7 and IL-15.^180–183 Despite these important differences, Tvm cells are often misclassified as central memory (Tcm) cells, especially when use of additional markers such as CD49d is omitted.¹⁷⁸ Although transcriptomic analysis suggests that Tvm cells from aged mice display a senescent-like phenotype, they are highly sensitive to IL-15 signals and do not exhibit the conventional proliferative defects associated with senescent cells.¹⁸⁴

Tvm cells from young mice respond more potently to TCR stimulation than Tvm cells from aged mice.¹⁸⁰ Tvm cells from aged mice display decreased proliferation and cytotoxic function upon T cell activation.^184,185 Paradoxically, aged Tvm cells possess a higher baseline and spare respiratory capacity (SRC), mitochondrial ETC complex IV expression and transcription of genes associated with mitochondrial biogenesis.^186,187 These metabolic properties of Tvm cells are dependent on IL-15 signaling.¹⁸⁷ Although Tvm cells accumulate in helminth and influenza infection and play a protective role in early stages of the infection in young individuals,^188–190 their role in immune and metabolic diseases in the aged remains unknown.

Exhausted-like GzmK⁺ T cells

Analysis of aged T cells at a single cell resolution recently identified age-associated CD8⁺ T cells (Taa) as a key hallmark of aging immunity and inflammaging.²⁹ Taa cells, characterized by the expression of granzyme K (GzmK), do not exist in the young, but become prevalent with age (Figure 3). Transcriptomic analyses and surface protein expression profiling suggest that GzmK⁺ Taa cells are distinct from CD8⁺CD28⁻, Temra and Tvm cells. Taa cells display features of exhausted T cells and are characterized by dual expression of PD-1 and Tox, two proteins critically involved in the maintenance of exhausted T cells. The transcriptome and epigenome of exhausted CD8⁺ T cells and Taa largely overlap, and share features including increased GzmK, Tox and Eomes and decreased Tbx21 and Tcf7 expression.²⁹ Galletti et al. also identified exhausted-like memory T (Tpex) cells which express GzmK, Tox and PD-1, and display key elements of exhausted T cells,¹⁹¹ yet it is unclear whether these cells expand with age. The precise mechanisms underlying Taa cell induction and maintenance are unclear, however adoptively transferred young lymphocytes fully convert into Tox⁺PD-1⁺CD49d⁺ Taa cells after 1 month in aged mice,²⁹ suggesting that environmental factors including cytokines and the metabolic niche may play a key role in Taa differentiation.

The concept of inflammaging was originally coined by Franceschi and colleagues who postulated that innate immune cells are central drivers of age-associated inflammation.¹⁹² Several studies have challenged this idea and evidence supporting a central role for aged T cells in inflammaging has emerged.^29,144 For example, GzmK⁺ Taa cells directly contribute to inflammaging.²⁹ In humans, GzmK expression in CD8⁺ T cells is positively correlated with increased levels of IL-6, TNF-α and IL-8, which are associated with inflammaging.²⁹ While the mechanisms driving Taa cell formation and inflammaging remain unclear, metabolic perturbations in T cells may directly promote inflammaging. T cell-specific ablation of TFAM, which plays a critical role in stabilization and replication of mitochondrial DNA, induced defective mitochondrial respiration and subsequently augmented pro-inflammatory type I helper (T_h1) T cell skewing and inflammaging, characterized by the precipitation of a premature aging phenotype and increased peripheral levels of IL-6, TNF-α and IFN-γ.¹⁴⁴ Furthermore, increasing the level of NAD⁺, a metabolic cofactor which plays an important role in mitochondrial function, through the use of NAD⁺ precursor nicotinamide riboside (NR), partially restored NAD⁺/NADH ratio and attenuated inflammaging induced by TFAM deficient T cells.¹⁴⁴ Thus, mitochondrial dysfunction in T cells may contribute to inflammaging.

Collectively, these studies highlight the phenotypic, functional and metabolic heterogeneity of age-related T cell subsets. At present, it is unknown whether these subsets are formed directly as a consequence of aging and are thereby primary in nature, or result from a compensatory process during aging, where maladaptive metabolic reprogramming plays a central role. Therefore, a deeper mechanistic understanding of the metabolic programs that govern the maintenance and functional responses of these cells may reveal novel metabolic vulnerabilities leading to dysfunctional T cell responses in aging.

Regulatory T cells

Regulatory T (Treg) cells expressing the transcription factor FoxP3 play a critical role in the maintenance of peripheral tolerance and prevention of autoimmunity.^193,194 Treg cells function by suppressing inappropriate immune responses directed against self-tissues, innocuous environmental antigens, as well as the commensal microbiome. Studies examining the contributions of the aging process on the maintenance and functional responses of Treg cells appear to be species and context dependent.¹⁹⁵ In a limited number of studies, the proportions and absolute numbers of Treg cells increased during the aging process while suppressive function was dependent on the inhibitory modality assessed.^195,196 For example, aged Treg cells controlled the proliferation of effector T cells in a classical in vitro Treg suppression assay, whereas they exhibited a marked deficit in their capacity to suppress IL-17 production from T helper 17 (T_h17) cells in an experimental mouse colitis in vivo.^197,198 Interestingly, decreased control of IL-17 production was associated with attenuated STAT3 activation in aged Treg cells, which is required for Treg cells to regulate T_h17 responses.

Recent work suggests that Treg cells exhibit a greater vulnerability towards aging as compared to effector T cells, and that aged Treg cells fail to protect mice from colitis and irradiation-induced aging.¹⁹⁹ Mechanistically, aged Treg cells were shown to downregulate the E3 ligase DDB1- and CUL4-associated factor 1 (DCAF1), which resulted in unbridled ROS production leading to Treg cell aging and aberrant inflammation.¹⁹⁹ These findings suggest that DCAF1 mediated restraint of ROS is a critical checkpoint that governs the susceptibility of Treg cells to senescence. Furthermore, global knockout of the deglycase DJ-1, an oxidative stress sensor which regulates ROS levels, resulted in further dysregulation of pyruvate dehydrogenase (PDH) activity and impairment of OXPHOS in aged Treg cells. These metabolic changes more profoundly attenuated the homeostatic proliferation and activation of Treg cells in aged as compared to young mice.²⁰⁰ Collectively these results suggest that the metabolic vulnerabilities of aged Treg cells contribute to disruption of Treg cell homeostasis and peripheral tolerance during the aging process.

VII. Anti-tumor immunity & Aging

The concept that the immune system can detect and effectively eliminate nascent tumor cells has been around for more than a century. In the 1860s, the German physicians F. Fehleisen and W. Busch observed that cancer patients who developed subsequent skin infections exhibited significant regression of tumors.^201,202 The administration of heat-killed bacteria into cancer patients was pioneered by the American surgeon W. B. Coley, who is often regarded as the father of cancer immunotherapy.^203,204 “Coley’s toxin,” the immune-modulatory extract developed by W. B. Coley, was shown to achieve clinical responses in a number of malignancies. Although significant progress has been made since these early observations, the role of T cells in cancer progression and/or elimination has remained underappreciated until recently. We now appreciate that T cells are indispensable for the surveillance and elimination of tumors, and numerous therapeutic strategies have been developed to capitalize on tumor-specific T cell responses.^205–207 Yet, experimental studies investigating the anti-tumor activity of T cells are almost exclusively conducted in young individuals. Thus, it remains unclear if and how age-associated T cell dysfunction contributes to anti-tumor immunity.

A connection between tumor growth and T cell-mediated responses in the aged has been underappreciated perhaps due to conflicting reports. Elderly patients bearing neoplasms display delayed hypersensitivity responses to DNCB, suggesting that T lymphocyte function may decline with age.¹⁶ Additional clinical studies have similarly reported correlations between accelerated tumor growth in the elderly with functional and numerical T cell deficits.^14,208 Subcutaneous injection of carcinogens such as 3-methylcholanthrene in aged C57/BL6 mice resulted in increased numbers of tumor-bearing mice characterized by decreased tumor induction periods compared to young mice.²⁰⁹ However, other early studies suggested that control of tumor growth may be increased in aged experimental models.^210–212 Today, the consequences of aging on anti-tumor T cell immunity and tumor control remain controversial. Although some studies describe defective anti-tumor immunity and poor tumor control in aged mice,¹²⁰ others demonstrate intact or even improved control of tumor growth in the aged setting.^213–215 Numerous factors can account for these differences, including the mouse strain, diet, sex, microbiome and type of tumor model used. Collectively, these reports provide impetus for a deeper investigation into the effects of aging on anti-tumor immunity.

Age-associated metabolic defects in anti-tumor T cells

The importance of metabolic checkpoints in the regulation of anti-tumor T cell immunity is now well appreciated.²¹⁶ Depletion of essential nutrients, hypoxia and production of immune-suppressive metabolites all play major roles in subverting T cell responses against tumors.^216,217 In addition to a metabolically hostile environment, there may be age-associated metabolic deficits in anti-tumor T cells. For example, defective 1-carbon metabolism in aged anti-tumor T cells contributed to higher tumor growth in mice bearing MC38 tumors.¹²⁰ Moreover, age-associated accumulation of CerS6/C14 ceramide induced T cell mitophagy and mitochondrial dysfunction, leading to impaired anti-tumor T cell response against B16 melanoma.¹¹⁹ In both studies, age-associated mitochondrial dysfunction upon TCR stimulation impaired tumor-specific naïve T cell priming, effector response and survival. Beyond T cell-intrinsic regulation, aging is also associated with alterations in host metabolism and the tumor microenvironment.²¹⁸ Understanding how aging changes the metabolic landscape in the tumor microenvironment may lead to a better understanding of how anti-tumor T cells are regulated with age. Notably, naïve CD8⁺ T cells represent a minor population within the total T cell pool in aged hosts; the immunologic and metabolic properties of CD8⁺CD28⁻ T cells, Temra cells, Tvm cells and Taa cells in the context of tumor immunity also require further investigation.

VIII. Checkpoint blockade & Aging

T cell responses are tightly controlled by multiple tolerogenic mechanisms which prevent collateral damage to healthy tissues and cells;^219–222 the tolerogenic mechanisms involved in the regulation of auto-reactive T cells are exploited by tumor cells to impair anti-tumor T cell immunity. In particular, the expression of inhibitory receptors, also known as “immune checkpoints,” is critical for the regulation of anti-tumor T cells.²²³

Following TCR engagement, T cells upregulate the inhibitory molecule CTLA-4, a structural homologue of the costimulatory receptor CD28, which binds to B7–1/B7–2 on APCs with higher affinity than CD28. This interaction promotes competitive inhibition of co-stimulatory signaling and leads to attenuated T cell activation. Removing this “brake pedal” through the use of anti-CTLA-4 blocking antibodies improved anti-tumor T cell responses, and eradication of established tumors in mice.²²⁴ This seminal finding became the foundation for the field of immune checkpoint blockade (ICB) with the subsequent demonstration that inhibition of CTLA-4 results in unprecedented long-term survival in a subset of patients with metastatic melanoma.²²⁵ The programmed death 1/programmed death ligand 1 (PD-1/PD-L1) pathway represents another cornerstone of immune regulation.^226,227 PD-1 is also rapidly upregulated in activated T cells and engages with its ligands PD-L1/PD-L2, which can be expressed on non-hematopoietic cells including tumor cells, to dampen TCR signaling and T cell activation. Moreover, PD-1 signals are critically involved in the maintenance of T cell exhaustion.²²⁷ Blockade of PD-1/PD-L1 can lead to the functional re-invigoration of a subset of exhausted CD8⁺ T cells ultimately resulting in control of tumor growth.²²⁸ PD-1 pathway inhibitors have been FDA-approved for the treatment of >20 types of cancer.

Although age is a major risk factor for developing cancer and patients 65 years of age or older represent up to 60% of newly diagnosed cases, few studies have addressed the functional consequences of an aged immune system on the efficacy of cancer immunotherapies including ICB. Of particular importance is the question of how age dependent metabolic alterations of immune cell populations, including metabolic reprogramming of aged anti-tumor T cells, impact the activity of ICB agents. Despite the unprecedented success of ICB, most patients do not respond to monotherapy, and many develop resistance following initial response.²²⁹ Combination therapy approaches may be key to enhancing anti-tumor activity and durability of response in older patients unresponsive to ICB. In the following section we survey preclinical studies examining the efficacy of ICB in tumors in aged mice, and summarize insights generated from clinical trials assessing the activity of ICB therapies in aged human subjects.

Pre-clinical studies

Aged mice harbor higher proportions of CD44⁺ effector-memory T cells characterized by high expression levels of PD-1 and its ligands PD-L1/PD-L2.^135,230 These PD-1⁺ T cells have limited proliferative potential, a diminished capacity to produce effector cytokines like IFN-γ, and do not exhibit durable functional re-invigoration following PD-1/PD-L1 blockade (Figure 4).¹³⁵ Expression of PD-L1/PD-L2 is also markedly increased in dendritic cell (DC) populations from aged mice including classical and myeloid DC subsets (Figure 4).^135,231 Several studies have documented comparable or even superior activity of PD-1/PD-L1 inhibitors in aged as compared to young tumor bearing mice.^232–234 In comparison to young mice bearing BSC9AJ2 melanoma tumors, which were resistant to the effects of anti-PD-1, aged mice exhibited greater sensitivity towards PD-1 blockade.²³³ In this study, aged mice exhibited decreased frequency of FoxP3⁺ regulatory T cells, an increased CD8⁺:Treg cell ratio and elevated numbers of IFN-γ and TNF-α cytokine producing CD8⁺ T cells.²³³ Consistent with these results, Treg cell depletion in young mice sensitized tumors to PD-1 inhibition.²³³ These results are consistent with observations that Treg cell depletion improves efficacy in young but not aged hosts with B16-tumors.²³⁵ In a subcutaneous model of oral cancer, treatment of aged mice with anti-PD-1/PD-L1 mAbs elicited superior anti-tumor activity compared to that of young mice.²³⁶ These studies support the notion that ICB efficacy is intact in certain aged tumor model systems.

Figure 4: Age-associated changes in anti-tumor T cell immunity and response to ICB.

Aging results in systemic remodeling of primary, secondary lymphoid tissues, and the tumor microenvironment, all of which have significant implications for anti-tumor T cell immunity and ICB. Increased myelopoiesis plays a central role in shaping the immunosuppressive environment in secondary lymphoid and tumor tissue, whereas thymic involution and defective T cell development limit TCR diversity and naïve T cells required for anti-tumor immunity. The literature suggests defective T cell priming and anti-tumor T cell responses occur with age. Furthermore, aging is associated with an increased pool of PD-1⁺ T cells and expression of PD-L1/PD-L2 on APCs. However, the overall response to ICB remains unclear, and may be context-dependent.

In contrast to the aforementioned reports, studies have also found that aging may subvert the activity of ICB therapy. Single agent anti-PD-L1 or anti-CTLA-4 therapy failed to control tumor growth in aged B16 tumor-bearing mice.^232,235 Coupling anti-PD-L1 therapy with Treg depletion enhanced the anti-tumor response in young but not aged B16 tumor-bearing mice.^232,235 In contrast, depletion of Tregs in aged mice restored protective anti-tumor immune responses in a lymphoma model.²³⁷ Expression of PD-1 and its ligands PD-L1/L2 is altered by aging in an organ- and tissue-specific manner and may account for differential ICB efficacy profiles in young vs. aged B16 tumor-bearing mice.²³¹ In murine models of triple negative breast cancer, aged animals exhibited impaired responsiveness to anti-PD-L1 or anti-CTLA-4 monotherapy which was driven by defective type 1 IFN signaling related to inadequate innate immune cell priming.²³⁸ Resistance to anti-PD-1 therapy in MC38 tumor-bearing aged mice was shown to result from ineffective generation of pre-effector-like T cells that rely on 1-carbon metabolism for adequate priming and effector differentiation,¹²⁰ consistent with previous findings in CD4⁺ T cells.⁵⁷ Taken together, these investigations reveal divergent outcomes to ICB in preclinical studies and suggest that responsiveness to ICB is context dependent and may differ according to tumor model and treatment.

It is important to note that no studies to date have examined how the metabolic niche and PD-1 signaling regulates myeloid cell development and function in aged tumor bearing mice. In mice 8–12 weeks of age, PD-1 blockade decreased the accumulation of immunosuppressive MDSCs and increased the systemic output of effector myeloid cells, including inflammatory macrophages and DCs.²³⁹ Dual MDSC and Treg cell depletion elicited superior anti-tumor activity in aged mice and was characterized by elevated IFN-γ production in T cells.²³⁵ Because aging is associated with major alterations in hematopoiesis including a marked bias towards the generation of myeloid progenitor HSCs (Figure 4), it is of great interest to determine whether the metabolic environment and PD-1 signaling have a significant role in immunosuppressive myeloid cell formation and maintenance during anti-tumor immune responses in aged mice. Lastly, a comprehensive understanding of the bioenergetic adaptations enforced by a hostile TME in which chronic TCR and PD-1 signals are integrated to shape unique metabolic vulnerabilities of T cells is needed to understand the metabolic determinants of effective anti-tumor immunity and the discrete metabolic pathway utilized by aged T cell subsets.

Clinical studies

Although a clear understanding of how aging impacts ICB responsiveness in cancer patients is precluded by the limited number of clinical studies, most available data support the premise that ICB efficacy is intact in elderly patients.^240–244 A recent meta-analysis exploring the efficacy of ICB based on age (younger vs. 65 years or older), spanning 17 randomized control trials in a variety of malignancies, determined that the PD-1/PD-L1 inhibitors pembrolizumab and atezolizumab had comparable efficacy in young adults and patients ≥65 years.²⁴⁰ In a separate meta-analysis of ICB combination therapies (either with other ICBs, chemotherapy, targeted therapy or radiotherapy) in NSCLC, efficacy was comparable between younger patients and patients with a cut-off age of 65 years.²⁴² Similar results have been reported by several groups.^243–245 Consistent with these findings, a higher prevalence of biomarkers associated with responsiveness to ICB, including tumor mutational burden (TMB), is observed in elderly patients.²⁴⁶ Further investigation is required to understand the differences observed in human and murine studies, and whether metabolic deficits of aged T cells provide an additional resistance barrier for ICB therapy.

IX. Intervention strategies to bolster aged T cells

As research continues to unravel novel strategies to boost immunity in older people, the question of how aging impacts anti-tumor immunity and responsiveness to ICB therapy will be further investigated. Given the lack of clinical and mechanistic insight linking aging and the response to immunotherapy in the elderly, an important future direction will be the identification of therapeutic interventions that can synergize with cancer immunotherapy, specifically in the aged setting. Several therapeutic approaches have been employed to overcome distinct features of the aging T cell,⁵ as will be briefly discussed below.

Decelerating thymic involution

A number of therapeutic strategies have been implemented to delay age-related thymic involution and increase thymic output. IL-22 produced by innate lymphoid cells can promote the proliferation and survival of thymic epithelial cells (TEC), ultimately driving thymic regeneration following radiation-induced damage of the thymus.²⁴⁷ Proliferation and differentiation of TECs in response to thymic injury resulting from total body irradiation is dependent on production of the fibroblast growth factor 7 (FGF7).^248,249 IL-7, which is produced by TECs and critically supports thymopoiesis, is decreased in the aged thymus but can restore defective thymic output when directed to the aged mouse thymus.^250,251 Heightened physical activity is associated with increased plasma IL-7 levels and decreased levels of IL-6, which promotes thymic atrophy and can induce thymic degeneration when administered to mice.^252–254 In addition to cytokines, targeting metabolic pathways driving thymic involution may contribute to the development of novel therapeutic strategies aimed at rejuvenating thymic output. However, this will require a better understanding of the metabolic mechanisms during thymic atrophy.

Overcoming age-associated T cell dysfunction

Metabolic interventions, including inhibition of mTOR can enhance the response to influenza vaccination in elderly human subjects and rejuvenate hematopoietic stem cell (HSC) function in aged mice.^255,256 Low dose inhibition of the mTOR complex TORC1 decreased the incidence of infections in the elderly and augmented anti-viral immunity.²⁵⁷ A series of nationwide clinical trials are currently underway to examine the benefits of the anti-hyperglycemic agent metformin, which inhibits components upstream of mTOR including mitochondrial complex 1 and AMP-activated protein kinase (AMPK), in preventing or delaying age-associated morbidities.²⁵⁸ Furthermore, a recent study demonstrated that targeted deletion of TFAM in T lymphocytes results in unabated inflammation and the precipitation of a premature aging phenotype, which were reversed upon TNF-α neutralization or supplementation with the NAD⁺ precursor nicotinamide riboside.¹⁴⁴ This finding is consistent with previous work showing that NAD⁺ is critically important for mitochondrial function and that reversing age-associated decline of NAD⁺ is protective against age-related morbidities.^259–261 Lastly, there are ongoing efforts to target senescent immune cells which serve as central drivers of the aging process. For example, use of the “senolytic” agent ABT-737, which preferentially eliminates senescent cells, alleviated the senescent cell burden and age-related morbidities seen in perforin-deficient mice.¹⁴⁵ The literature suggests that the use of senolytic agents, which were shown to improve the health and life span of aged mice,^262,263 may rejuvenate the overall immune system and suppress age-associated morbidities. In the context of tumor immunity, recent studies have investigated and targeted senescent T cells for improving response to immune checkpoint blockade.^264,265 Collectively, these studies suggest targeting mTOR, TNF-α, NAD⁺ and/or senescent cells may reduce the systemic effects of aging and have significant implications for improving anti-tumor immunity in the aged setting.

X. Concluding remarks

Over the past decade, much has been learned about the hallmarks of T cell aging and the contributions of the aging process to maladaptive T cell immunity.⁵ Despite this progress, our understanding of fundamental mechanisms responsible for the emergence of these hallmarks remains largely obscure, as does our knowledge of the interplay between features of the aging T cell in the context of an aged microenvironment. Therefore, it is paramount to not only understand the primary causes of distinct hallmarks of T cell aging but how their interactions shape and perpetuate age-associated T cell dysfunction. Amongst the hallmarks of aging, metabolic dysfunction has emerged as a central orchestrator of the aging process in T cells. Given that metabolic dysregulation is inherently interconnected with and may causally drive other hallmarks of T cell aging, it is critical understand the mechanisms governing metabolic dysregulation in the aged T cell. Identifying the metabolic mechanisms driving physiological T cell aging may reveal novel points of interception for rejuvenation of the aged T cell. Whether such rejuvenation is necessary and sufficient to reverse multi-organ aging and senescence remains yet to be elucidated. Likewise further work is needed to determine whether rejuvenation of aged and/or senescent T cells can strengthen anti-tumor immunity and response to ICB.²⁶⁶

In this review, we have highlighted fundamental principles governing the metabolic reprogramming of T cells and metabolic alterations that occur in the aged. Although much has been learnt about the phenotypic and functional characteristics of age-associated T cell subsets, a comprehensive understanding of the metabolic programs associated with their formation, maintenance and functional contribution to disease processes is only beginning to emerge. Similarly, the implications of age-associated metabolic reprogramming in cancer immunotherapy, including the metabolic pathways used and vulnerabilities of aged anti-tumor T cells, are not yet clear. The advent of high-dimensional approaches, including genetic screens and perturbation approaches, provide means to identify novel regulators of metabolic dysfunction in aged T cells and selective metabolic modulation strategies to improve T cell function in the elderly. Developing a better understanding of how aging reprograms T cell metabolism and immunity should contribute to rational design of cancer immunotherapies for the elderly.