DEREGULATED NUTRIENT-SENSING
The nutrient-sensing network is highly conserved in evolution. It includes extracellular ligands, such as insulins and IGFs, the receptor tyrosine kinases with which they interact, as well as intracellular signaling cascades. These cascades involve the PI3K-AKT and the Ras-MEK-ERK pathways, as well as transcription factors, including FOXOs and E26 factors, which transactivate genes involved in diverse cellular processes. The mechanistic target of rapamycin (MTOR) complex-1 (MTORC1) responds to nutrients, including glucose and amino acids, and to stressors such as hypoxia and low energy to modulate the activity of numerous proteins including transcription factors such as SREBP and TFEB. This network is a central regulator of cellular activity, including autophagy, mRNA and ribosome biogenesis, protein synthesis, glucose, nucleotide and lipid metabolism, mitochondrial biogenesis, and proteasomal activity. Network activity responds to nutrition and stress status by activating anabolism if nutrients are present and stress is low or by inducing cellular defense pathways in response to stress and nutrient-shortage. There is extensive intracellular crosstalk and feedback within the network, and between it and other intracellular signaling pathways. Genetically reduced activity of components of the nutrient-sensing network can increase lifespan and healthspan in diverse animal models178,179 (Table 1). Moreover, genetic association studies in humans have implicated the FOXO3 transcription factor180 and genetic variants encoding components of the network in human longevity.178 Epigenetic age is also associated with nutrient-sensing in human cells.181 In youth, activity of this signaling network thus functions to promote beneficial anabolic processes, but during adulthood, it acquires pro-aging properties (Figure 4).
The somatotrophic axis—the first one historically implicated in the control of aging—is a growth-stimulatory cascade that, at its apex, involves growth hormone (GH) produced by the hypophysis. GH acts on the GH receptor of hepatocytes to stimulate the secretion of IGFs, in particular IGF1, which promotes growth and development via the IGF1R to stimulate trophic signals through activation of PI3K-AKT and the MTORC1 network.182 In multiple model organisms, spontaneous or engineered mutations of this pathway enhance lifespan and retard facets of age-associated deterioration (Table 1). Innate defects in the somatotrophic axis cause dwarfism, but inhibition of this axis from early adulthood has beneficial effects on organismal health (Figure 4).
Another signaling pathway involved in nutrient-sensing relies on the receptor tyrosine kinase ALK (Figure 4), which, in mice, is induced in the hypothalamus by feeding183 and responds to the ligands augmentor a and b (Auga and Augb).184 In Drosophila, knockdown of ALK decreases triglyceride levels and the expression of several insulin-like peptides, whereas genetic or pharmacological inhibition of ALK extends healthspan and lifespan, mostly in females.183 In mice, body-wide or hypothalamus-specific deletion of ALK, as well as double knockout of Auga and Augb, promotes resistance against diet-induced obesity, and in humans, a loss-of-function mutation of ALK is associated with leanness.183,184 Hence, this pathway may offer additional targets for interventions on metabolic aging.
Drugs targeting diseases such as cancer and metabolic disease often engage the nutrient-sensing network, thus such drugs are candidates for repurposing as geroprotectors. Rapamycin and rapalogs, which disrupt the MTORC1 complex, have proved to extend lifespan in model organisms even with treatment starting late in adulthood.185 In mice, rapamycin can increase diverse aspects of health, although it exacerbates some age-related traits such as cataract, and it is protective in models of neurodegenerative and other age-related diseases.
Elderly humans are susceptible to viral respiratory infections. Pre-treatment with MTORC1 inhibitors increased the immune response of elderly volunteers to immunization against influenza186 and reduced viral respiratory infections in the ensuing winter,187 thus pointing to a potential strategy for reverting age-related immunosenescence.
Mechanisms
In humans, IGF1 peaks during the second decade of life but declines with aging. Inhibition of the GH/IGF1 pathway in adult or late life extends lifespan in model organisms, including mice. 48 Inhibition of cardiac IGF1R by expression of a dominant negative p110a isoform of PI3K increases maximum lifespan of male mice and improves heart function in aged mice.188 Moreover, enzymatic inhibition of IGF1R with tyrosine kinase inhibitors improves anticancer immunosurveillance requiring autophagy induction in malignant cells.189 Long-term administration of an anti-IGF1R antibody enhances the longevity of female (but not male) mice, although reducing inflammation and tumor development. These findings suggest that the IGF1/IGF1R signaling axis may constitute a target for anti-aging interventions. In favor for this conjecture, in elderly women (R95 years), as well as in a mixed population of older adults (mean age 76 years), low IGF1 levels correlate with a low probability of cognitive impairment and death.190 Moreover, in a large cohort from the UK Biobank, significant positive correlations were noted between the hazard associated with high IGF-1 and age for dementia, diabetes, vascular disease, osteoporosis, and overall mortality.191 In centenarians, the concentrations of IGF1BP2 and IGFBP6 are elevated. 192 Future will tell whether yet-to-be-developed antibodies or small molecules that selectively inhibit IGF1R signaling without affecting other receptor tyrosine kinases (and in particular the insulin receptor) might be used for the pulsatile inhibition of the somatotropic axis to achieve health benefits with acceptable side effects.
Effects of nutrition
Diet is one of the most practical targets for interventions into human aging. Mechanistically, overnutrition: (1) triggers intracellular nutrient sensors, such as MTORC1 (activated by leucine and other amino acids), and the acetyltransferase EP300 (activated by acetyl coenzyme A); (2) inhibits sensors that detect nutrient scarcity, such as AMP-activated kinase (AMPK) and the deacetylases SIRT1 and SIRT3 (which respond to NAD+ ); and (3) abolishes catabolic reactions (glycogenolysis, proteolysis for gluconeogenesis, and lipolysis coupled to ketogenesis) with consequent suppression of adaptive cellular stress responses, including autophagy, antioxidant defense, and DNA repair. Conversely, fasting and dietary restriction inhibit MTORC1 and EP300; activate AMPK, SIRT1, and SIRT3; and stimulate adaptive cellular stress responses as they suppress the somatotrophic axis and extend longevity in multiple model organisms including primates.193
Nutrient sensors constitute targets for potential longevity drugs (Figure 4), but health benefits and extended lifespan might also be achieved by dietary restrictions. Mechanistically, this is possible via reduction of overall caloric intake, manipulation of the dietary composition,194,195 or time-restricted feeding.196 Dietary restriction regimens are particularly successful in extending lifespan in male C57BL/6J mice, if the animals are completely deprived from nutrients during daytime.49 However, dietary restriction regimens do not extend lifespan in all mouse strains, supporting the contention that they must be adapted to the genetic makeup of each individual.197 In humans, clinical assays based on dietary restriction are complicated by poor compliance, yet suggest positive effects on immunity and inflammation. 50
Intermittent fasting (e.g., 1 day without nutrients, followed by 1 day of ad libitum feeding) can avoid long-term weight loss induced by caloric restriction, yet increases lifespan in mice195 and improves biomarkers of health in clinical trials.198,199 Life time extension of a similar intermittent fasting regimen in flies has been attributed to the nighttime-specific upregulation of autophagy-stimulatory genes,200 but this has not yet been investigated in mammals. Rapamycin-induced longevity extension (which in flies partially depends on autophagy induction) can be obtained by constant-long term exposure, as well as by intermittent regimens,201 suggesting that pulsatile inhibition of this axis is sufficient to obtain the benefits of lifespan extension. The optimal interval for such intermittent treatments has not yet been determined for clinical use, although partial caloric restriction for 4–7 days every 3–4 weeks may be sufficient to improve metabolic syndrome and anticancer immunosurveillance. 202
Another potentially beneficial regimen is ketogenic diet, which is a low-carbohydrate, high-fat, and adequate protein diet. Both fasting and ketogenic diet increase the production of ketone bodies (in particular 3-hydroxybutyrate), which are synthesized from acetyl coenzyme A in the liver in an autophagy-dependent fashion, can reach millimolar concentrations in the plasma and replace glucose as an essential fuel, for instance, for the maintenance of brain function.203 Permanent but not cyclic administration of 3-hydroxybutyrate in the drinking water increases lifespan and healthspan in mice.51 This strongly suggests that this ketone body mediates some of the beneficial effects of ketogenic diet. Mechanistically, 3-hydroxybutyrate induces vasodilatation and activates immune responses acting on GTP protein coupled receptor 109A,203 whereas it directly inhibits the NLRP3 inflammasome,204 indicating a potential pleiotropic mode of action.
MITOCHONDRIAL DYSFUNCTION
Mitochondria are not only the powerhouses of the cell but also constitute latent triggers of inflammation (when reactive oxygen species [ROS] or mtDNA leak out of the organelle causing activation of inflammasomes or cytosolic DNA sensors, respectively) and cell death (when activators of caspases, nucleases, or other lethal enzymes are released from the intermembrane space).146 With aging, mitochondrial function deteriorates due to multiple intertwined mechanisms including the accumulation of mtDNA mutations, deficient proteostasis leading to the destabilization of respiratory chain complexes, reduced turnover of the organelle, and changes in mitochondrial dynamics. This situation compromises the contribution of mitochondria to cellular bioenergetics, enhances the production of ROS, and may trigger accidental permeabilization of mitochondrial membranes causing inflammation and cell death.182 Logically, the function of mitochondria is primordial for the maintenance of health, and its progressive deterioration contributes to the aging phenotype (Figure 4).
Mitochondrial function and longevity
Healthspan-extending interventions can stimulate the function of mitochondria. For instance, placebo-controlled trials have revealed positive effects of L-carnitine supplementation on both pre-frail subjects and elderly men57 (Table 1). The effect is possibly mediated by counteracting age-related declining L-carnitine levels which may limit fatty acid oxidation by mitochondria. 205 Paradoxically, in model organisms, lifespan can be improved by compromising mitochondrial function, which induces a hormetic response (‘‘mitohormesis’’), provided that this inhibition is partial and occurs early during development. In C. elegans, partial inhibition of mitochondrial protein synthesis or import enhances lifespan through a mechanism involving the mitochondrial UPR (UPRmt).206 In Drosophila, muscle-specific knockdown of complex I subunit NDUFS1/ND75 extends longevity in an UPRmt-dependent fashion.207 Mild inhibition of mitochondrial ATP synthesis with TPP-thiazole can improve metabolic health in aging mice, reducing visceral fat and of senescent cells: (1) the transcriptional derepression of endogenous retroviruses, most notably LINE-1, which causes cytosolic leakage of double-stranded DNA and activates the cGAS/STING and TLR pathways;136 (2) the mitochondrial overproduction of ROS; and (3) the perturbation of the autophagy-lysosomal system leading to an expansion of lysosomal content that facilitates the histochemical detection of lysosomal senescence-associated beta-galactosidase (SABG).227
SASP is highly heterogeneous, depending on the cell type-specific activation of innate immunity signaling pathways (cGAS/ STING, TLRs, and NLRPs), mTORC1, and transcription factors (NF-kB, CBPs, GATA4, and others). SASP usually has simultaneous and partially conflicting consequences on the microenvironment: (1) to recruit and activate immune cells through the secretion of chemokines (CCL2, CXCL2, and CXCL3) and cytokines (IL-1b, IL-2, IL-6, and IL-8); (2) to suppress the immune system through the secretion of TGF-b; (3) to trigger fibroblast activation and collagen deposition through pro-fibrotic factors (TGF-b, IL-11, and PAI1); (4) to remodel the ECM through the secretion of matrix metalloproteases; (5) to trigger the activation and proliferation of progenitor cells through the secretion of growth factors (EGF and PDGF); and (6) to trigger paracrine senescence in neighboring cells (TGF-b, TNF-a, and IL-8). In many diseases, the net effect of SASP is chronic inflammation and progressive fibrosis.228
Although there is not a single unequivocal marker of cellular senescence, this process can be identified by the co-existence of a combination of features that, together, are specific and provide a molecular definition to the phenomenon:216 (1) lysosomal expansion, detectable by SABG; (2) upregulation of CDK inhibitors, particularly p16 and/or p21; (3) loss of LMNB1 from the nuclear envelope; (4) loss of the chromatin component HMGB1 from the nucleus and its extracellular release as an alarmin; (5) heterochromatic foci, visualized as HP1g nuclear foci or SAHFs; (6) high levels of ROS; (7) exacerbated DNA damage, visualized as gH2AX nuclear foci; and (8) high levels of SASP factors, notably IL-6, TGF-b, PAI1, and others.
Given the association between cellular senescence and multiple pathologies, the question arises about the biological purpose of such a cellular response. Cellular senescence is a potent tumor suppressor mechanism, but mounting evidence has linked cellular senescence to tissue repair processes in which senescent cells promote localized fibrosis and the recruitment of immune cells that then remove damaged and senescent cells. In this regard, tissue repair can be considered a two-step process: cellular senescence followed by immune recruitment and immune clearance of senescence (Figure 5A). In this scenario, senescence is a temporally restricted response that programs its self-elimination with a beneficial outcome.229 The pathological consequences of senescence only become visible when the second step of immune clearance is not achieved, and the accumulation of senescent cells and the SASP effects on the tissue microenvironment eventually result in fibrosis.
Figure 5. Cellular senescence and stem cell exhaustion
(A) Cellular senescence usually promotes tissue repair after injury and protects the organism from oncogenic damage. This is achieved in two steps:
(1) establishment of senescence and (2) recruitment of immune cells that will eliminate the senescent cells, thereby promoting tissue repair. If any of these steps fails, the organism is prone to develop diseases.
Senolytics
The strong association between cellular senescence and multiple pathologies has spurred the searchfor small chemical compounds that selectively kill senescent cells and that are referred to as ‘‘senolytics. 230’’ Of note, senolysis (elimination of senescent cells) is very different from the cancellation of the senescence response, which can result, for example, from mutation of p16 or p21. Senolysis does not prevent the execution of senescence but rather recapitulates the natural immune clearance of senescent cells (Figure 5A). In support of this, mice subjected to long-term genetic-induced or pharmacologically induced senolysis present extended longevity without increased cancer incidence or signs of defective tissue repair.59,58
The number of senolytic therapies is still limited, but some have been extensively used in preclinical models of disease, as exemplified by navitoclax, dual treatment with dasatinib and quercetin (D/Q), fisetin, cardiac glycosides, and others.221 The survival and apoptotic resistance of senescent cells strongly depends on the BCL2 family of proteins, specially BCLXL, but also BCL2 and BCLW. This renders senescent cells highly vulnerable to navitoclax, which targets these three proteins.231 Navitoclax has been evaluated in clinical trials for antitumor activity and it is expected that this drug (or derivatives lacking toxicity on platelets) will enter clinical trials for senescence-associated diseases.232 Other potential senolytic treatments such as D/Q230 and fisetin60 are approved for human use and are being tested in various clinical trials for multiple indications. The mechanistic basis for their action remains unclear. Dasatinib is a promiscuous kinase inhibitor, and quercetin and fisetin are natural flavonoids with multiple targets. D/Q has been tested in clinical trials with promising results in the case of lung and kidney fibrosis.62,61 Cardiac glycosides inhibit the plasma membrane Na+ /K+ -ATPase present in all cells causing a cationic imbalance and lowering the intracellular pH.233 The mechanism of senolysis by cardiac glycosides is likely connected to the vulnerability of senescent cells to low intracellular pH. Thus, chemical inhibition of glutaminase deprives cells of a mechanism to counteract low pH and results in senolysis.234 All the above-discussed senolytic compounds exert therapeutic activity in a wide range of murine disease models associated with senescence. Senolysis can also be achieved by immunological approaches that target proteins appearing on the surface of senescent cells. In particular, antibodies directed against the glycoprotein NMB (GPNMB)235 and CAR T cells directed against the receptor uPAR236 attenuate senescence-associated disease models in mice.
In summary, cellular senescence is an important response to stress and damage that, in normal physiology, is followed by immune clearance, but that upon aging or chronic damage fails to be eliminated by immune mechanisms and hence is pathogenic due to the abundant secretion of pro-inflammatory and profibrotic factors. Therapeutic strategies aimed at killing senescent cells have been extensively explored in animal models and are now in clinical trials (Table 1).
STEM CELL EXHAUSTION
Aging is associated with reduced tissue renewal at steady state, as well as with impaired tissue repair upon injury, with each organ having its own strategy for renewal and repair.237 For example, in skeletal muscle, one single-cell type, the satellite cell, is placed at the apex of a unipotent and unidirectional hierarchy, both for renewal and repair. In skin epidermis, which is characterized by high renewal and exposure to injury, there are multiple stem cell niches, particularly in association to the hair follicles, each one generating its progeny and territory. However, upon injury, multiple cells can acquire stem cell properties and subvert territorial boundaries. Other organs like liver, lung, or pancreas exhibit rather low renewal rates under normal conditions, contrasting with the acquisition of stem cell properties including proliferation and multipotency by different cell types (Figure 5B). Indeed, tissue repair is believed to rely to a large extent on injury-induced cellular de-differentiation and plasticity. For example, in the intestine, brain, and lung, injury induces dedifferentiation of non-stem cells, which reactivates normally silent embryonic and stemness transcription programs, thus acquiring the plasticity needed for tissue repair.238–240 Injuryinduced plasticity (and its progressive loss with aging) may be more relevant for aging than the plasticity of resident stem cells under normal homeostatic conditions. Stem and progenitor cells are all subject to the same hallmarks of aging as are cells without stem potential, and for this reason, we do not discuss here the abundant literature about the impact of each hallmark of aging on stem cell function. Instead, we will focus on a general strategy to counter the decline of stem cell function with aging based on the concept of ‘‘cellular reprogramming.’’ This process is thought to act in a cell-autonomous manner on multiple cell types; however, its impact on stem and progenitor cells is considered of higher relevance because of its long-term impact.
Figure 5. Cellular senescence and stem cell exhaustion
(B) Stem cell exhaustion results from the loss of cellular plasticity required for tissue repair. Tissue repair requires a modified microenvironment through the secretion of cytokines (in part due to the senescence-associated secretory response), growth factors and modulators of the extracellular matrix (ECM) that favors the de-differentiation and plasticity of cells from different tissue compartments. These injury-induced plastic cells may acquire multipotent progenitor function. Transient expression of OSKM factors represses the transcription of cell identity programs causing global de-differentiation (OSKMon) and the acquisition of plasticity. For rejuvenation, the process must be interrupted at this point (OSKMoff) to allow cells to re-differentiate and to restore their original cell identities.
Rejuvenation of tissue repair by reprogramming Cellular reprogramming toward pluripotency consists in the conversion of adult somatic cells into embryonic pluripotent cells (known as induced pluripotent stem cells or iPSCs) by the concomitant action of four externally transduced transcription factors, namely, OCT4, SOX2, KLF4, and MYC (OSKM).241 The process of reprogramming usually requires several weeks during which cells first lose their differentiated phenotype by transcriptional repression of cell identity genes and subsequently transactivate pluripotency genes.242 Full reprogramming not only implies a change of cellular identity but also cellular rejuvenation, characterized by a number of aging features that are reset to the embryonic state, as indicated by p16 reduction,243 extension of telomeres,244 and resetting of the DNA methylation clock.245 Interestingly, rejuvenation occurs in a progressive fashion starting shortly after the initiation of de-differentiation.246 Indeed, it is possible to initiate reprogramming with OSKM, interrupt the process at an intermediate state, and allow cells to return to their original identity. This transient cellular perturbation, variously known as ‘‘partial,’’ ‘‘transient,’’ or ‘‘intermediate’’ reprogramming, is able to rejuvenate cellular markers of aging such as the DNA methylation clock, DNA damage, epigenetic patterns, and aging-associated changes in the transcriptome, both in vitro and in vivo. 63,64,70,246,247 Therefore, it can be proposed that the processes of de-differentiation and rejuvenation are coupled. Specifically, de-differentiation implies the erasure of epigenetic and transcriptional programs, and this may also erase aging-associated alternations. Upon interruption of partial reprogramming, cells re-stablish their original epigenetic and transcriptional status in a process of re-differentiation that, interestingly, does not re-stablish the erased aging-associated changes and therefore resets the epigenome and transcriptome to a younger state.
Transient reprogramming in mice confers repair capacity to old tissues so that a subsequent damage is repaired as efficiently as in young individuals. This increased repair capacity has been shown for models of tissue damage in the endocrine pancreas,63 skeletal muscle,63 nerve fibers,70 eye,70 skin,64 heart,65 and liver.66 Also, tissue dysfunctions characteristic of natural aging, such as reduced visual acuity70 and the loss of adult neurogenesis in the hippocampus and long-term memory,67 can be partially reversed by transient reprogramming. There are a few instances in which transient reprogramming is beneficial also during the process of tissue repair (and not only prior to the injury). This is the case for traumatic brain injury68 and skin wound healing.69 Finally, it should be mentioned that the lifespan of progeroid mice can be extended by transient reprogramming,63 although extension of longevity by OSKM has not yet been reported for wild-type mice.
Partial reprogramming recapitulates features of natural tissue repair (Figure 5B). In both cases, cells undergo a transient process of de-differentiation, acquisition of embryonic and progenitor features, and subsequent re-differentiation. Thus, de- and re-differentiation could explain tissue rejuvenation, in line with the observation that transient de-differentiation of myocytes, followed by their re-differentiation, induces rejuvenation of the transcriptome.248 The natural process of tissue repair may imply some degree of cellular rejuvenation, in accord with the finding that the epigenetic methylation clock accelerates soon after tissue injury and partially reverses during tissue repair.249 Moreover, tissue damage reportedly creates a tissue microenvironment that is highly permissive for IL-6-driven reprogramming. 250 Finally, cyclic expression of the transcription factor FOXM1 extends the longevity of progeroid mice and wild-type mice.251 Although the detailed mechanism is still unexplored, FOXM1 is induced in the kidney upon injury and participates in triggering de-differentiation and proliferation of tubular epithelial cells during the repair process.252 Thus, several features of natural tissue repair and artificial reprogramming may converge, perhaps allowing refinement of strategies for restoring repair capacity in aging tissues.
ALTERED INTERCELLULAR COMMUNICATION
Aging is coupled to progressive alterations in intercellular communication that increase the noise in the system and compromise homeostatic and hormetic regulation. Thus, aging involves deficiencies in neural, neuroendocrine, and hormonal signaling pathways, including the adrenergic, dopaminergic, and insulin/IGF1-based and renin-angiotensin systems, as well as sex hormones commensurate with the loss of reproductive functions. 182,253 Although the primary causes of such alterations are cell intrinsic, as this is particularly well documented for the SASP, these derangements in intercellular communication ultimately sum up to a hallmark on its own that bridges the cell-intrinsic hallmarks to meta-cellular hallmarks including the chronification of inflammatory reactions coupled to the decline of immunosurveillance against pathogens and premalignant cells, as well as the alterations in the bidirectional communication between human genome and microbiome, which finally results in dysbiosis. A number of studies in this regard have focused on the search for blood-borne systemic factors with pro-aging or prolongevity properties, the role of diverse communication systems between cells, and the evaluation of the functional relevance of ECM disruption during aging.
Pro-aging blood-borne factors
A single transfusion of old blood induces features of aging in young mice within a few days,72 and the simple dilution of the blood of old mice with saline buffer containing 5% albumin induces rejuvenation in multiple tissues,71 indicating the existence of circulating factors that favor the aging process. Among the pro-aging blood-borne factors, the chemokine CCL11/eotaxin and the inflammation related protein b2-microglobulin reduce neurogenesis,254,255 IL-6 and TGF-b impair hematopoiesis,256 and the complement factor C1q compromises muscle repair.257 Theoretically, the neutralization of these factors might have potent anti-aging effects. Indeed, several among the aforementioned factors are secreted in the context of SASP and may be co-responsible for the phenomenon of ‘‘contagious’’ aging, which also involves extracellular vesicles.258 Thus, socalled ‘‘senomorphics’’ might be used to repress SASP and slow down aging.
Anti-aging blood-borne factors
Soluble factors present in the blood of young mice effectively restore renewal and repair capacity in old mice259 (Table 1). Heterochronic parabiosis experiments followed by extensive singlecell transcriptomics have confirmed the capacity of young blood to rejuvenate multiple tissues74 and to restore age-associated reduction in general gene expression, in particular that of mitochondrial genes involved in the electron transport chain.75 The chemokine CCL3/MIP-1a acts as a rejuvenating factor for hematopoietic stem and progenitor cells;74 the metalloproteinase inhibitor TIMP2 has been implicated in rejuvenating the hippocampus;73 the anti-inflammatory interleukin IL-37 (which declines in monocytes from aged humans) improves increased endurance exercise and ameliorates whole-body metabolism in old mice;76 the cytokine GDF11 rejuvenates some tissues, such as muscle, brain, and endocrine pancreas, although it impairs the function and repair of other tissues due to its pro-fibrotic side effects;77 and finally, mice with transgene-enforced VEGF overexpression exhibit enhanced liver and muscle repair, improved general health and an extension in average longevity by ~40%.78
Long-range and short-range communication systems
The central nervous system controls multiple facets of aging affecting peripheral organs, explaining how brain-specific gene manipulations like overexpression of SIRT1, UCP1, or knockout of IKBKB and TRPV1 can enhance mouse longevity (Table 1). The precise mechanisms of these long-range activities are yet to be determined.260 Of note, intercellular communication also involves the interaction among short-lived extracellular molecules (such as ROS, nitric oxide, nucleic acids, prostaglandins, and other lipophilic molecules), soluble factors that are released from various tissues including white adipose tissue (adipokines), brown adipose tissue (baptokines), heart (cardiokines), liver (hepatokines) and skeletal muscles (myokines, including exerkines produced in response to exercise), cell-bound ligands, and receptors on other cells (as exemplified by IL-1a that can remain cell-bound), as well as direct cell-to-cell interactions mediated by tight junctions or gap junctions. All these communication systems may be altered during aging and hence are being scrutinized for their potential pro- and anti-aging properties.258
Extracellular matrix
Aging causes numerous damages in the long-lived protein components of the ECM, including AGEs, carbonylation and carbamylation, elastin fragmentation, and collagen crosslinking,261 thus leading to tissue fibrosis (fibroaging).262 This deleterious process is in part due to the excessive release of TGF-b and other growth factors, and the nuclear translocation of TAZ and YAP transcription factors, which act as mechanotransducers and trigger the expression of pro-fibrotic genes such as transglutaminase-2, lysyl oxidase (LOX), and LOX-like enzymes. 262 ECM stiffness also affects the function of senescent cells, which in turn secrete matrix metalloproteases that amplify the damage of the ECM,263 and proteolytically generate damage-associated molecular patterns to activate pro-senescent, pro-fibrotic, and pro-inflammatory pathways.262 The increasing stiffness of the aging matrix may also favor WNT signaling to induce fibroblast activation and expression of pro-fibrotic genes. 264 This pathway exhibits extensive crosstalk with other pro-fibrotic pathways, such as NOTCH, RAS, TGF-b/SMAD, and hedgehog/GLI, thereby demonstrating the complexity and interconnections of mechanisms underlying the development of age-linked fibrosis.262 Of note, mechanical change caused by matrix stiffness is sufficient to cause age-related loss of function of oligodendrocyte progenitor cells in a process mediated by the mechanoresponsive ion channel PIEZO1.265
Several studies have provided causal evidence for the contribution of ECM stiffness to aging and have also suggested approaches for improving healthy aging (Table 1). In vivo inhibition of Piezo1 using AAV vectors results in rejuvenation of the oligodendrocyte progenitors in the brain of old mice.265 Genetic inactivation of YAP/TAZ in stromal cells causes accelerated aging, although sustaining YAP function rejuvenates old cells and prevents the emergence of aging features by controlling cGAS-STING signaling.79 Moreover, mice engineered to produce collagenase-resistant type I collagen (Col1a1r/r) exhibit vascular cell senescence, accelerated aging, and shortened lifespan.266 The importance of collagen for human longevity has been reinforced by the discovery of rare variants in COL25A1—encoding a brain-specific collagen—that may have a protective role against Alzheimer’s disease.267 Moreover, ECM prepared from young human fibroblasts induces a youthful state in aged senescent cells.80 ECM compounds such as chondroitin sulfate and hyaluronic acid restore the age-related decline of collagen and increase lifespan in nematodes.268 Conversely, ectopic expression of human hyaluronidase TMEM2 promotes resistance to ER stress and extends lifespan in C. elegans through changes in p38/ERK MAPK signaling.269 In mice, deletion of chondroitin 6-sulfotransferase results in an abnormal ECM in the brain, early memory loss, and accelerated brain aging, whereas overexpression of this enzyme improved memory in old mice.81 Retrospective analyses indicate that oral intake of glucosamine/chondroitin sulfate leads to a reduction in all-cause mortality in humans.270 However, there is no prospective proof thus far that such a prolongevity effect would be mediated through an amelioration of the ECM.
CHRONIC INFLAMMATION
Inflammation increases during aging (‘‘in-flammaging’’) with systemic manifestations, as well as with pathological local phenotypes including arteriosclerosis, neuroinflammation, osteoarthritis, and intervertebral discal degeneration. Accordingly, the circulating concentrations of inflammatory cytokines and biomarkers (such as CRP) increase with aging. Elevated IL-6 levels in plasma constitute a predictive biomarker of allcause mortality in aging human populations. 271 In association with enhanced inflammation, immune function declines, a phenomenon that can be captured by high-dimensional monitoring of myeloid and lymphoid cells in the blood from patients and from mouse tissues.272 For example, a population of age-associated T cells—termed Taa cells—is composed of exhausted memory cells that mediate pro-inflammatory effects via granzyme K. Shifts in T cell populations entail the hyperfunction of pro-inflammatory TH1 and TH17 cells, defective immunosurveillance (with a negative impact on the elimination of virus-infected, malignant or senescent cells), loss of self-tolerance (with a consequent age-associated increase in autoimmune diseases), and reduced maintenance and repair of biological barriers, altogether favoring systemic inflammation273 (Figure 6A).
Figure 6. Derangement of supracellular functions
Altered intercellular communication bridges the cell-intrinsic hallmarks to meta-cellular hallmarks including the chronic inflammation, and the alterations in the crosstalk between human genome and microbiome, which finally result in dysbiosis.
(A) Chronic inflammation during aging occurs as a consequence of multiple derangements that stem from all the other hallmarks. Several representative examples of anti-inflammatory interventions with positive effects on healthspan and lifespan are shown in the right part of the figure.
(B) Dysbiosis contributes to multiple pathological conditions associated with aging. The human gut microbiota significantly changes during aging, finally leading to a general decrease in ecological diversity. The main features of the mechanisms underlying these microbiota changes and some examples of interventions on the gut microbiota composition which can promote healthy aging are shown in the lower part of the right panel. CVDs, cardiovascular diseases; SCFAs, short-chain fatty acids.
Links between inflammation and other aging hallmarks
Inflammaging occurs as a result of multiple derangements that stem from all the other hallmarks. For example, inflammation is triggered by the translocation of nuclear and mtDNA, into the cytosol where it stimulates pro-inflammatory DNA sensors, especially when autophagy is ineffective and hence unable to intercept ectopic DNA.4 Genomic instability favors clonal hematopoiesis of indeterminate potential (CHIP), with the expansion of myeloid cells that often bear a pro-inflammatory phenotype, driving for instance cardiovascular aging.274 Intriguingly, the most frequent CHIP-associated mutations affect the epigenetic modifiers DNMT3 (which methylates cytosine residues in DNA) and TET2 (which catalyzes the oxidation of methylcytosine to 5-hydroxymethylcytosine). Mechanistically, CHIP affecting TET2 enhances IL-1b and IL-6 production by myeloid cells and stimulates cardiovascular disease (CVD), which is attenuated among individuals bearing a loss-of-function mutation in the IL-6 receptor or treated with an IL-1b neutralizing antibody. 275
Overexpression of pro-inflammatory proteins can be secondary to epigenetic dysregulation, deficient proteostasis, or disabled autophagy. Excessive trophic signals resulting in activation of the GH/IGF1/PI3K/AKT/mTORC1 axis trigger inflammation. In addition, inflammation is favored by the SASP secondary to the accrual of senescent cells, as well as by the accumulation of extracellular debris and infectious pathogens, which are not cleared due to senescence, and by exhaustion of myeloid and lymphoid cells. This latter phenomenon involves age-associated thymic involution, abrogating thymopoiesis with the consequent rarefaction of the T cell repertoire and the inability to mount efficient immune responses against novel antigens.276 Of note, thymopoiesis is improved by CR in humans, and a CR-downregulated gene coding for platelet activation factor acetyl hydrolase A2 group VII (PLA2G7) can be knocked out in mice to combat thymic atrophy.50 Finally, inflammaging is also exacerbated by perturbations of circadian rhythms and by intestinal barrier dysfunction.277
Anti-inflammatory, anti-aging interventions
Although systemic inflammation is mechanistically linked to all the aforementioned age-associated alterations, inflammation constitutes a hallmark on its own. Indeed, specific manipulations of the inflammatory and immune system can accelerate or decelerate the aging process across different organ systems. For example, a T cell-specific defect in the mitochondrial transcription factor A (TFAM) is sufficient to drive cardiovascular, cognitive, metabolic, and physical aging coupled to an increase in circulating cytokines. The TNF-a inhibitor etanercept partially reversed this phenotype.82 Heterozygous deletion of the DNA repair protein ERCC1 in hematopoietic cells from mice is sufficient to induce immunosenescence and aging of non-lymphoid organs, as well as numerous signs of organ damage coupled to reduced lifespan. This phenotype was alleviated by the senolytic fisetin.278 These results support the idea that aging of the immune system may drive organismal aging. Of note, adoptive transfer of TFAM-null T cells, young ERCC-deficient splenocytes, or aged wild-type splenocytes into young mice induced senescence, whereas the transfer of young immune cells into ERCC-deficient mice attenuated senescence, pointing to the capacity of immune cells to modulate organismal aging in both positive and negative terms.82,278
There are multiple examples of broad healthspan and lifespanexpanding effects of anti-inflammatory treatments (Figure 6A; Table 1). Thus, blockade of TNF-a prevents sarcopenia in mice and improves cognition in aging rats.83,84 Blockade of the common type 1 interferon receptor (IFNAR1) reverses the accumulation of monocytes in the aging mouse lung.279 Knockout of the prostaglandin E2 receptor EP2 in myeloid cells or treatment of aged mice with pharmacological EP2 inhibitors ameliorates cognition. 85 Knockout of the inflammasome protein NLRP3 improves metabolic biomarkers, glucose tolerance, cognition, and motor performance and extends mouse longevity.86 Pharmacological inhibitors of NLRP3 or of its downstream enzyme caspase-1 have encouraging preclinical effects on normal and accelerated aging models.280 Most importantly, inhibition of the caspase-1 product IL-1b with canakinumab exemplifies an anti-aging treatment applicable to patients. The phase 3 clinical trial CANTOS evaluated the capacity of canakinumab to prevent recurrent CVD in patients with a history of myocardial infarction and signs of pronounced inflammation. Beyond meeting the primary endpoint of the trial, canakinumab reduced the incidence of diabetes and hypertension, as well as the incidence of, and mortality from, lung cancer.87 Finally, although long-term use of non-steroidal anti-inflammatory agents such as aspirin may have positive effects on human health—in particular with respect to the prevention of CVD and gastrointestinal cancers—a large phase 3 clinical trial in which aspirin was administered to over 70-year-old subjects yielded negative results.281 Hence, further studies will be necessary to explore the value of prophylactic treatments with aspirin at a younger age to combine aspirin with other medications or to replace aspirin by less toxic anti-inflammatory drugs.
DYSBIOSIS
Over recent years, the gut microbiome has emerged as a key factor in multiple physiological processes such as nutrient digestion and absorption, protection against pathogens, and production of essential metabolites including vitamins, amino acid derivatives, secondary bile acids, and short-chain fatty acids (SCFAs). The intestinal microbiota also signals to the peripheral and central nervous systems and other distant organs and strongly impacts on the overall maintenance of host health.146 Disruption of this bacteria-host bidirectional communication results in dysbiosis and contributes to a variety of pathological conditions, such as obesity, type 2 diabetes, ulcerative colitis, neurological disorders, CVDs, and cancer.282 The progress in this field has generated an enormous interest in exploring gut mi crobiota alterations in aging (Figure 6B).
Microbiota alterations in aging
The microbial community within the intestinal tract is highly variable among individuals as a consequence of host genetic variants (ethnicity), dietary factors, and lifestyle habits (culture), as well as environmental conditions (geography), which makes difficult to unveil the relationships between microbiota and pleiotropic age-associated disease manifestations. Nonetheless, some meta-analyses have revealed microbiota-disease associations that have been validated across distinct pathologies283 and countries.284,285 Studies in both humans and animal models have provided valuable information on clinical, epidemiological, sociological, and molecular aspects that underlie the complex effects of an aged microbiome on human health and disease.286 Once bacterial diversity is established during childhood, it remains relatively stable during adulthood. However, the architecture and activity of this bacterial community undergoes gradual changes during aging, finally leading to a general decrease in ecological diversity. Thus, several studies conducted on centenarian populations showed a reduction in core abundant taxa, such as Bacteroides and Roseburia, but also an increase in several genera such as Bifidobacterium and Akkermansia, which appear to have prolongevity effects.287
These studies have been extended by recent analysis of the gut microbiome and phenotypic data from over 9,000 individuals of three independent cohorts spanning 18–101 years of age.288 Of note, individual gut microbiomes become increasingly more unique to each individual with age, and this uniqueness is associated with well-known microbial metabolites involved in immune regulation, inflammation, and aging. In older age, healthy participants show continued drift toward a unique microbial composition, whereas this drift is reduced or absent in individuals in worse health. The identified microbiome pattern of healthy aging is characterized by a depletion of core taxa, such as Bacteroides, present across most humans. Moreover, in individuals approaching extreme age, retention of high Bacteroides levels and a low gut microbiome uniqueness measure are significantly associated with decreased survival. However, findings in microbiota from centenarians and supercentenarians are not always concordant with those derived from elderly populations. The ELDERMET study reported an increased dominance of the core genera Bacteroides, Alistipes, and Parabacteroides in old individuals compared with younger controls. This study also identified age-related shifts in gut microbiota composition linked to frailty, cognition, depression, and inflammation.289 Another study revealed age-related trajectories of the microbiota shared across populations of different ethnicities, as well as a common age-related decrease in sex-dependent differences in gut microbiota. Of note, older adults exhibit higher abundances of several health-promoting bacterial species, including Akkermansia. 290 These results suggest that agerelated physiological changes, beyond dietary changes and lifestyle of older adults, may have profound effects on the human gut microbiota.
The heterogeneity of findings in all these studies indicates that there may be multiple gut microbiome trajectories of aging. However, there is an interesting convergence in plasma concentrations of microbiota-produced amino acid derivatives. These metabolites include indoles derived from gut bacterial degradation of tryptophan, and several fermentation products of phenylalanine/tyrosine, such as p-cresolsulfate, phenylacetylglutamine, and p-cresol glucuronide. This finding is consistent with data from the ELDERMET cohort showing that fecal concentrations of p-cresol correlate with increased frailty and may contribute to age-associated decline in this population. Conversely, plasma concentrations of certain indole metabolites correlate with improved fitness in older adults. Indole metabolites increase healthspan and lifespan in mice, at least in part, by attenuation of inflammatory responses through binding of the arylhydrocarbon receptor.93
Further metabolomics and functional analysis of the gut microbiome of centenarians have shown its enrichment in some particular bacteria, such as Alistipes putredinis and Odoribacter splanchnicus. Some of these bacterial species are capable of generating unique secondary bile acids, including isoallo-lithocholic, which exerts potent antimicrobial effects against grampositive multidrug-resistant pathogens such as Clostridioides difficile and Enterococcus faecium. 291 Thus, specific bile acid metabolism may be involved in reducing the risk of pathobiont infection and contribute to intestinal homeostasis, thereby decreasing the susceptibility to age-associated chronic diseases.
Fecal microbiota transplantation and aging
Multiomics studies in pathological aging have revealed that two different mouse models of progeria exhibit intestinal dysbiosis mainly characterized by an increase in the abundance of Proteobacteria and Cyanobacteria and a decrease in levels of Verrucomicrobia. Consistent with these findings, human progeria patients with HGPS or NGPS also show intestinal dysbiosis, whereas long-lived humans exhibit a substantial reduction in Proteobacteria and a significant increase in Verrucomicrobia.88 The causal implications of these changes were demonstrated in vivo by fecal microbiota transplantation (FMT). FMT from wild type to progeroid mice recipients enhanced healthspan and lifespan in both accelerated-aging models, whereas administration of the verrucomicrobium Akkermansia muciniphila was also sufficient to obtain such effects. Conversely, FMT from progeroid donors to wild-type recipients induced detrimental metabolic alterations. Restoration of secondary bile acids and other metabolites depleted in progeroid mice phenocopied the beneficial effects of reestablishing a healthy microbiome88 (Table 1).
FMT also revealed the causative role of gut dysbiosis in the chronic systemic inflammation and the decline in adaptive immunity associated with aging and age-related diseases. Transfer of the gut microbiota from old mice to young germfree mice triggered inflammatory responses characterized by enhanced CD4+ T cell differentiation in spleen, upregulation of inflammatory cytokines, and increased circulation of inflammatory factors of bacterial origin.292 FMT also provided evidence for the implication or the gut microbiota in the maintenance of brain health and immunity during aging.90 Microbiota from young mice donors reversed aging-associated differences in hippocampal metabolites and brain immunity and ameliorated ageassociated impairments in cognitive behavior when transplanted into an aged host. These works open the possibility of manipulating the gut microbiota with pre-, pro-, and post-biotics to rejuvenate the immune system and the aging brain. Heterochronic fecal transfers confirmed the causal link between age-dependent changes in microbial composition and a decline in the function of the host immune system.92 Indeed, the defective germinal center reaction in Peyer’s patches of aged mice can be rescued by FMT from younger animals without affecting germinal center reactions in peripheral lymph nodes. Finally, FMT from young donor mice improves ovarian function and fertility in aged mice. These beneficial effects are associated with an improvement in the immune microenvironment of aged ovaries, with decreased macrophages and macrophage-derived multinucleated giant cells, reduced levels of pro-inflammatory IFNg, and increased abundance of the anti-inflammatory cytokine IL-4.91
To be continued
