Targeting Senescent Cells for a Healthier Aging: Challenges and Opportunities

Abstract Aging is a physiological decline in both structural homeostasis and functional integrity, progressively affecting organismal health. A major hallmark of aging is the accumulation of senescent cells, which have entered a state of irreversible cell cycle arrest after experiencing inherent or environmental stresses. Although cellular senescence is essential in several physiological events, it plays a detrimental role in a large array of age‐related pathologies. Recent biomedical advances in specifically targeting senescent cells to improve healthy aging, or alternatively, postpone natural aging and age‐related diseases, a strategy termed senotherapy, have attracted substantial interest in both scientific and medical communities. Challenges for aging research are highlighted and potential avenues that can be leveraged for therapeutic interventions to control aging and age‐related disorders in the current era of precision medicine.

In the time course of organismal aging, senescent cells are increasingly accumulated in most, if not all, types of tissues and organs. Originally identified as a tumor suppressor mechanism restricting the expansion of damaged or neoplastic cells, cellular senescence plays indispensable roles in the pathogenesis and exacerbation of a wide spectrum of chronic diseases during aging, defining cellular senescence as a representative example of evolutionary antagonistic pleiotropy. [ 9 , 10 ] In the tissue microenvironment (TME), senescent cells mount complex autocrine and paracrine responses by continuously secreting a myriad of extracellular factors including growth factors, cytokines, chemokines, metalloproteases and various extracellular matrix remodeling elements, a feature branded as senescence‐associated secretory phenotype (SASP), or sometimes, senescence‐associated secretome. [ 11 , 12 , 13 ] While these cells are mostly deleterious in later life stages, they do contribute to a subset of physiological events in the organismal lifespan, including a) tissue repair and wound healing via secretion of certain SASP factors, including but not limited to platelet‐derived growth factor AA (PDGF‐AA); [ 14 ] b) tissue remodeling and regeneration in the context of injury and ageing, such as creation of a permissive TME for in vivo reprogramming via IL‐6 production; [ 15 , 16 , 17 ] c) enhancement of therapeutic vulnerabilities via production of pro‐angiogenic SASP factors such as VEGF, PDGFA/B, and FGF2 to allow tumor vascularization, eventually promoting drug delivery and efficacy of cytotoxic chemotherapy and sensitizing tumors to PD‐L1‐mediated checkpoint blockade; [ 18 ] and d) embryonic development by promoting morphogenesis of tissues such as the mesonephros and endolymphatic sac of human embryos. [ 19 , 20 ] Paradoxically, senescent cells can engage immune surveillance to promote their self‐elimination, a process implicating the recruitment of phagocytic immune cells and mobilization of their nearby progenitors. [ 21 , 22 ] However, in the case of persistent damage and/or during natural aging, physiologically autonomous clearance of senescent cells is compromised and dysfunctional cells are accumulated, together generating a chronic pro‐inflammatory microenvironment and contributing to various pathological conditions across the lifespan. [ 23 , 24 ] Experimentally described by Hayflick and colleagues in 1960s, cellular senescent was first reported in their studies of human diploid fibroblasts which experienced a limited number of consecutive divisions under culture conditions, a feature later attributed to telomeric attrition. [ 25 , 26 ] To date, numerous studies suggest that cellular senescence can occur after exposure of cells to different types of stimuli including DNA damage, epigenetic alteration, proteostasis loss, mitochondrial dysfunction, oncogenic activation, oxidative stress, and even high‐fat diet (HFD) ( Table 1 ). Table 1 A representative list of stimuli and inducers responsible for cellular senescence Stimulus or inducer Treatment example or induction route In vivo consequence Telomeric attrition Suppressors of telomerase activity (e.g., SYUIQ‐5,3′‐azido‐3′‐deoxythymidine, pyridostatin) Aging; cancer [ 123 , 124 ] Genotoxic agents Inducers of DNA replication stress (e.g., bromodeoxyuridine, hydroxyurea); Genotoxic drugs including DNA topoisomerase inhibitors (e.g., etoposide, doxorubicin, mitoxantrone), DNA crosslinkers (e.g., mitomycin C, cisplatin) and drugs with complex effects (e.g., actinomycin D, methotrexate bleomycin) Cancer regression accompanied by off‐target or side effects [ 32 , 110 , 125 , 126 ] Deregulated nutrient sensing Perception of intracellular and/or extracellular nutrient signals (amino acids, glucose, NAD+) by signaling mediated by insulin IGF1, mTOR, or AMPK Aging‐associated changes [ 127 , 128 ] Ionizing irradiation X‐ray, γ ‐radiation, UV light Cancer regression accompanied by side effects [ 125 , 126 ] Oncogene activation HRas, KRas, NRas, BRAF Cancer progression or suppression [ 129 , 130 , 131 ] Loss of tumor suppressors p53, PTEN Tumor inhibition; cancer progression [ 132 , 133 ] Oxidative stress Inducers of reactive oxygen species (e.g., paraquat, hydrogen peroxide) Aging [ 134 ] Mitochondrial dysfunction Decline of mitochondrial malix enzyme 1 (ME1) and malix enzyme 2 (ME2), reduced NAD + level or decreased NAD + /NADH ratio Aging [ 135 , 136 ] Loss of proteostasis ER stress, mTOR activation, UPR events Aging [ 137 , 138 , 139 ] Suppression by cyclin‐dependent kinase inhibitors Upregulation of CDKIs such as p16 INK4a /p19 ARF /p21 CIP1 (downstream of p53) (e.g., nutlin 3a); senescence‐inducing drugs (e.g., abemaciclib, palbociclib, ribociclib) Tumor suppression; cancer progression [ 140 , 141 ] Cytokines TGF‐ β and analogs Aging [ 142 , 143 ] Activators of protein kinase C PEP005, PEP008, TPA/PMA, Aging [ 144 ] Epigenetic modifications DNA methyltransferase sup

Beyond the Bcl‐2 family members, there are alternative molecules that are of prominent value as targets for senolytics. HSP90 enhances the survival of senescent cells of some tissue types by stabilizing the cytoplasmic level of p‐Akt, and HSP90 suppressors have recently been found to be senolytic for some senescent cell subpopulations. [ 47 ] Similar to “D+Q” and Fisetin, the HSP90 inhibitor 17‐DMAG drives clearance of senescent cells, extend of healthspan of animals, and postpone the onset of certain age‐related conditions. Expression of FOXO4, a transcription factor, is upregulated in senescent cells and prevents apoptosis of senescent cells via sequestration of p53 in the nucleus, while a synthetic FOXO4 peptide (namely FOXO4‐DRI) is able to selectively remove senescent cells from the TME niche through p53‐dependent apoptosis. [ 48 ] The cyclin‐dependent kinase inhibitor p21 CIP1 , a downstream molecule of p53, protects senescent cells of certain tissue origins against death by dampening signal transduction events that engage JNK and caspase in a setting of persistent DNA damage. [ 49 ] Besides causing apoptosis in non‐senescent cells, Nutlins, a group of small‐molecule antagonists of MDM2, can induce both cellular senescence and apoptosis, especially in the case of glioblastoma multiforme, a most common and invasive form of primary brain tumor. [ 50 ] Specifically, some Nutlins such as Nutlin‐3a, are cis‐imidazoline analogs that can cause apoptosis across a number of cell types (including non‐senescent cell types) by targeting MDM2‐p53 interactions, resulting in abrogated glioblastoma growth or a dampened SASP. [ 51 , 52 ] Intra‐articular administration of a Nutlin, UBX0101, depletes senescent cells in mouse synovium and articular cartilage induced by surgical damage to the knee (US patent 10130628). [ 33 ] At least in mice, injection of the Nutlin UBX0101 into the knee alleviates post‐surgical osteoarthritis, reducing pain while partially restoring cartilage. However, effects of UBX0101 on pain in a clinical trial of this agent administered into knee joints of patients with femoro‐tibial osteoarthritis were mixed ( NCT03513016 ), and there is hitherto no clear demonstration of senescent cell clearance or radiological improvement in the joints of these patients. [ 53 ] A recent study indicated that ouabain, a cardiac glycoside (CG), can act as an excellent candidate for senolytics, as senescent cells are efficiently eliminated via apoptosis mediated partially by engagement of the pro‐apoptotic NOXA, a member of the Bcl‐2 superfamily. [ 54 ] Indeed, senescent cells tend to display a depolarized plasma membrane and harbor increased concentrations of H + , features that make them more sensitive to the stimuli of CGs. [ 55 ] Data from in vivo studies support that CGs are also effective as senolytics in alleviating senescence‐induced lung fibrosis. [ 55 ] However, CGs can have serious side effects, including cardiac arrhythmias, heart block, and confusion, particularly in the elderly. [ 56 ] The curcumin analog, EF24, is a potent senolytic for several types of human senescent cells, such as fibroblasts, endothelial cells, and renal epithelial cells, mainly by reducing the expression of anti‐apoptotic factors Bcl‐xL and Mcl‐1. [ 57 ] The alkaloid piperlongumine, which derives from long peppers and several medicinal plants, can eliminate senescent human WI38 fibroblasts that arise after exposure to ionizing radiation, replicative exhaustion or oncogenic RAS G12V activation. [ 58 ] However, pretreatment with Q‐VD‐OPh, a pan‐caspase inhibitor, markedly abrogates piperlongumine‐induced apoptosis of senescent cells, implying involvement of caspase signaling in the senolytic activity of piperlongumine. [ 58 ] Expression of the antioxidant protein, oxidation resistance 1 (OXR1), which governs the synthesis of many antioxidant enzymes, is elevated in senescent WI38 fibroblasts, but piperlongumine can bind OXR1 directly and enhances its degradation through ubiquitin‐proteasome, thereby forcibly causing cell death [ 59 ] ( Figure 2 A ). Figure 2 Emerging strategies that showed efficacy in targeting cellular senescence. At late or advanced stages of life, senescent cells progressively accumulate in multiple tissue types and substantially promote the development of multiple age‐related diseases. Increasing lines of studies suggest that targeting specific fundamental mechanisms of aging may be more effective than treating each chronic pathology individually to restrain the impacts of aging on healthspan and lifespan. Pilot trials have been initiated to therapeutically target cellular senescence, such as A) selectively removing senescent cells via induction of their death (senolytic treatment), and B) engagement of the immune surveillance (senescence‐directed immunotherapy), with the potential of each strategy in depleting senescent cells from tissue microenvironment demonstrated by recent investigations. In each mechanist

In contrast to aforementioned intrinsic pathway‐directed depletion of senescent cells, therapeutic approaches mediated by the immunity represent a different strategy that is attracting substantial attention. Senescent cells are immunogenic per se and are subject to immune surveillance. [ 22 , 83 ] A natural but distinct mechanism responsible for dampened inflammatory response during cellular senescence (particularly replicative senescence and therapy‐induced senescence) and aging in Spalax was recently discovered, providing a rationale for developing senolytic agents that act through the immune system. [ 84 ] Although senescent cells can chemo‐attract various immune cell types, a process promoting senescent cell clearance by the immune components, many studies have suggested that senescent cells can also circumvent innate and adaptive immune responses. Therapeutics enhancing immune‐potentiated clearance of senescent cells include but are not limited to vaccines and small molecule immune modulators. [ 85 ] Senescent cells including those induced by replicative exhaustion, oncogene activation and DNA‐damaging agents, display enhanced expression of ligands for NKG2D, a homodimeric C‐type lectin‐like receptor present on the surface of NK and cytotoxic T cells. [ 86 ] Thus, NKG2D ligands may empower immune‐mediated clearance of these cells. Mechanisms of senescent cell surveillance can be context‐dependent. Expression of the oncogene NRAS G12V induces senescence in hepatocytes, generating a combined phenotype of both innate and adaptive immune responses, whereby CD4+T cells cooperate with both monocytes and macrophages to clear senescent hepatocytes. [ 22 ] However, senescent cells can evade immune recognition via engagement of MMP‐dependent shedding of NKG2D ligands such as MICA/B and ULBP1‐3, thus disabling NKG2D‐mediated immune surveillance in a paracrine manner. [ 87 ] Although expressing more soluble MICA/B ligands than replicatively senescent or therapy‐induced senescent cells, oncogene RAS G12V ‐induced senescent cells evade NKG2D receptor‐mediated immune surveillance by expressing high levels of MMP1/3/10/12, a group of SASP factors that have protease activities. [ 87 ] Thus, immune evasion and persistence of RAS G12V senescent cells is mediated, at least in part, by active protease secretion and high levels of NKG2D Ligand shedding. Such coordinated immune editing also takes place in residual, drug‐resistant tumors from breast and prostate cancer patients after genotoxic chemotherapy, which frequently induces cellular senescence in damaged tissues, [ 87 ] exemplifying how senescent cells can elude immune surveillance and implying the necessity of senolytic strategies to enhance clearance of senescent cells arising after clinical treatment or accumulating in aged tissues. Alternatively, administration of polyI:C, an immune booster, helps promote elimination of senescent cells mediated by NK cells in fibrotic livers, facilitating resolution of fibrosis induced by acute tissue damage. [ 88 ] Nevertheless, in vivo treatment with a strong immune stimulator may likely overwhelm the immune system in an aged individual, such as the case of cytokine storm observed in patients developing severe COVID‐19 and manifesting acute respiratory distress syndrome, the leading cause of mortality among those affected in current global range. [ 89 , 90 ] Thus, clinically safer and more precise immune modulators need to be pursued for this purpose. Beyond immune‐boosting therapeutic strategies, an augmenting arsenal of tools including chimeric antigen receptor T cells, present the avenue to facilitate redirecting immune responses against senescent cells. [ 91 ] However, a set of highly selective senescence biomarkers is not yet available that can allow this strategy to be pursued. Senescent cells are partially protected from NK cell‐mediated cytotoxicity via upregulation of decoy receptor 2 (DcR2), a member of the tumor necrosis factor‐related apoptosis‐inducing ligand (TRAIL) receptor family. [ 92 ] DcR2 suppresses activation of death receptors 4 and 5 (DR4/5) caused by TRAIL, confining cell death to perforin‐ and granzyme‐mediated signaling. [ 93 ] Thus, searching for ways to promote these activities, such as shielding DcR2, might confer senescent cells with increased vulnerability to innate immune surveillance. Although technical advances regarding immune‐mediated senescent cell clearance are still preliminary and await continued investigation, there appears charming potential for harnessing the immune system to target senescent cells. [ 28 ] Identification of specific and novel senescent cell surface markers might accelerate the development of immune‐associated senotherapies (Figure 2B ).

Aging a complex and dynamic process that drives progressive decline of tissue functionality, physiological integrity and its regenerative potential. Demographic data suggest that the aging rate of global populations continues to rise, a tendency resulting in the prevalence of various chronic pathologies and production of public health epidemics. [ 114 , 115 ] Although preclinical studies convincingly support that senescent cell elimination can ameliorate or even revert the pathological progress of multiple diseases, there are challenges that need to be well taken to ensure successful translation of interventional strategies. Senescent cells are increasingly abundant within tissues in the course of aging. Despite such a fact, cell senescence is essential in certain physiological aspects such as tumor suppression, tissue repair, and regeneration, wound healing, vascular remodeling and embryonic development in a damage‐independent manner. [ 14 , 15 , 18 , 20 , 116 , 117 ] Thus, unselectively manipulating cellular senescence during these processes may compromise physiological integrity and affect patient health. Although the effect of senolytics on individual physiological processes remains largely unexplored in the experimental context, current senolytics generally induce apoptosis of senescent cells upon a transient exposure, with their administration limited to a short period of time to avoid ensuing side‐effects. Further, the targeting specificity of senolytics represents a challenging issue. There are possible strategies to overcome this difficulty and optimize selectivity. First, a feasible solution is to develop pharmacologically active compounds and probes already tested in preclinical studies by exploiting synthetic lethal vulnerabilities, such as the drug delivery system taking advantage of the increased lysosomal β ‐galactosidase activity of senescent cells to develop agents including galacto‐oligosaccharides and the recently reported senescence‐specific killing compound 1 SSK1, wherein the acetyl group and β ‐gal‐responsive group improves cellular permeability and specificity. [ 46 , 118 ] Second, an alternative option for enhancing specificity is to modify the therapeutic or diagnostic agent that can be activated by external stimuli or enzymatic actions, as exemplified by the case of galactose‐based nanoparticles, encapsulation methods and probes utilizing β ‐galactosidase expression of senescent cells. [ 118 , 119 , 120 ] Third, expression of the SASP components by senescent cells might be employed to stimulate drug delivery systems and inactive pro‐senolytics or synthetic probes, for in vivo imaging purposes including magnetic resonance imaging and positron emission tomography. [ 27 ] Synthetically lethal compounds would be loaded into nanoparticles targeting senescent cells, thereby extensively widening their therapeutic windows, efforts requiring a more detailed understanding of the senescent cell‐specific signaling pathways or regulatory networks. For future clinical studies, there is still a major need to develop a set of novel, accurate and noninvasive senescence biomarkers for reliable assessment the burden of cellular senescence in the TME. Patient candidates for senotherapies will be appropriately evaluated according to their senescent burden, an essential step ensuring interventional necessity and clinical success. Given the heterogeneity of clinical patients and the complexity nature of aging and age‐related disorders, future interventions against cellular senescence will likely be context‐dependent, or more accurately, personalized. In response to the urgent and unmet clinical, healthcare, workforce, and economic needs of global aging populations who are suffering from aging‐related conditions and multimorbidity, the world health organization (WHO) has recently called for a public health attention based on the law of human rights. [ 121 ] To conduct a clinical trial, a corresponding disease classification code is necessary and can be adopted according to the WHO international classification of diseases guideline. [ 122 ] A comprehensive and systematic staging system would allow classification of aging individuals from the point of absence of organ and tissue senescence pathology and any appearance, features, and diagnostic criteria. Such a proposed framework is intended to be applied independently or combined with currently present classification codes in a complementary pattern, across disease diagnosis, prevention, and clinical intervention. [ 122 ] In the long run, the success of any clinical management to counter multimorbidity will be affected by the wish of an elderly individual to minimize its effects, and correspondingly, his/her compliance with preventative measures. For those willing, however, lifestyle improvement protocols and preventative interventions are handily available, with a line of emerging and promising regimens on the new horizon of precision medicine. Despite a