www.aging-us.com AGING 2019, Vol. 11, No. 16 Research Perspective In silico clinical trials for anti-aging therapies Javier A. Menendez1,2,*, Elisabet Cuyàs1,2,*, Núria Folguera-Blasco3, Sara Verdura1,2, Begoña Martin-Castillo4, Jorge Joven5, Tomás Alarcón6,7,8,9 1ProCURE (Program Against Cancer Therapeutic Resistance), Metabolism and Cancer Group, Catalan Institute of Oncology, Girona, Spain 2Girona Biomedical Research Institute (IDIBGI), Girona, Spain 3Quantitative Cell Biology Lab, The Francis Crick Institute, London, United Kingdom 4Unit of Clinical Research, Catalan Institute of Oncology, Girona, Spain 5Unitat de Recerca Biomèdica (URB-CRB), Hospital Universitari de Sant Joan, Institut d'Investigació Sanitària Pere Virgili, Reus, Spain 6ICREA, Barcelona, Spain 7Centre de Recerca Matemàtica (CRM), Barcelona, Spain 8Departament de Matemàtiques, Universitat Autònoma de Barcelona, Barcelona, Spain 9Barcelona Graduate School of Mathematics (BGSMath), Barcelona, Spain * Equal contribution Correspondence to: Javier A. Menendez, Tomás Alarcón; email: jmenendez@idibgi.org, talarcon@crm.cat Keywords: aging, cancer, inflammation, senolytics, biomathematics Received: February 21, 2019 Accepted: August 9, 2019 Published: August 24, 2019 Copyright: Menendez et al. This is an open‐access article distributed under the terms of the Creative Commons Attribution License (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. ABSTRACT Therapeutic strategies targeting the hallmarks of aging can be broadly grouped into four categories, namely systemic (blood) factors, metabolic manipulation (diet regimens and dietary restriction mimetics), suppression of cellular senescence (senolytics), and cellular reprogramming, which likely have common characteristics and mechanisms of action. In evaluating the potential synergism of combining such strategies, however, we should consider the possibility of constraining trade-off phenotypes such as impairment in wound healing and immune response, tissue dysfunction and tumorigenesis. Moreover, we are rapidly learning that the benefit/risk ratio of aging-targeted interventions largely depends on intra- and inter-individual variations of susceptibility to the healthspan-, resilience-, and/or lifespan-promoting effects of the interventions. Here, we exemplify how computationally-generated proxies of the efficacy of a given lifespan/healthspan-promoting approach can predict the impact of baseline epigenetic heterogeneity on the positive outcomes of ketogenic diet and mTOR inhibition as single or combined anti-aging strategies. We therefore propose that stochastic biomathematical modeling and computational simulation platforms should be developed as in silico strategies to accelerate the performance of clinical trials targeting human aging, and to provide personalized approaches and robust biomarkers of healthy aging at the individual-to-population levels. A growing aging population burdened by multiple in Western society [1]. Elderly individuals who are frail chronic diseases and geriatric syndromes has escalated are less able to cope with physiological stressors, and to become a major public health and medical challenge not only experience suboptimal health outcomes but www.aging-us.com 6591 AGING also contribute to rising healthcare costs [2,3]. Thus, the effects, while different mechanisms could be targeted in combined effects of increased life expectancy and the combination to enhance the effects. projected exponential growth of the aging population will critically reduce the capacity of our medical care The attractive possibility of achieving supra-additive systems to improve healthcare quality in the elderly. In outcomes when therapeutically addressing human aging this scenario, the traditional biomedical focus on single with combinatorial strategies should consider the diseases, which often reduces mortality without parallel existence of constraining trade-off phenotypes preventing or reversing the decline in overall health, (Figure 1). For example, because senescent cells are will result in an inadequate resolution of the societal known to have beneficial effects including facilitating burden of chronic illnesses and frailty [1-4]. tissue repair after injury and preventing tissue fibrosis, an unrestricted inhibition of senescence and/or A better understanding of how a limited number of inflammation with the use of senolytics could fundamental processes drives aging has led to the significantly impair the ability of the organism to emergence of candidate drugs and non-pharmacological undergo normal tissue repair and remodeling while strategies specifically targeting the so-called hallmarks promoting tissue-specific fibrosis [12-16]. Also, of aging [5-9]. It is safe to predict that the next 10 to 20 DR/DRM strategies can suppress senescence and pro- years will see a range of aging-related treatments inflammatory cytokine levels, potentially interfering clinically tested for their ability to lower all-cause with normal wound healing processes [17,18]. Chronic mortality by targeting complex multimorbidity (i.e., suppression of senescence, a primary mechanism of healthspan promoters), to boost the capacity of cells tumor suppression [19,20], and unrestricted cellular and tissues to respond or recover from an acute health reprogramming, might similarly lead to excessive stress (i.e., resilience promoters), or even to reverse distortion of the epigenome landscape, loss of cell functional decline via organismal rejuvenation (i.e., identity and tumorigenesis [21,22]. It therefore appears lifespan promoters). Although some of these drugs are that the disruption of entry-exit mechanisms and already in clinical use for other specific indications kinetics of endogenous injury-repair mechanisms is the (e.g., metformin and rapamycin) [10,11], the design of convergent trade-off of all the emerging anti-aging bona fide anti-aging clinical trials with the existing (and strategies. Accordingly, whereas transient epigenomic with a yet-to-be funded) pipeline of healthspan- remodeling in response to partial cellular reprogram- /resilience-/lifespan-promoting treatments will pose a ming phenomena is a rejuvenation strategy for erasing the hallmarks of aging at the molecular and cellular unique set of challenges, namely: a.) the occurrence of level [23-26], tissue repair impairment and tumori- intra- and inter-individual variability of response to genesis could also occur as the trade-off if the aging-targeting strategies with a consequent need for unrestricted epigenetic remodeling in response to aging- personalized treatments; and b.) the need for reasonable targeted strategies is not accompanied by self-repair of end points and robust biomarkers of healthy aging to injury or disease [27]. assess the response to aging-targeted interventions at the individual-to-population levels. Endogenous heterogeneity and individual-to- individual variability: two key regulators of the The trade-offs of aging-targeted strategies: the benefit/risk ratio of anti-aging strategies epigenetic nature of a convergent concern Beyond genetic factors, non-genetic stimuli such as The majority of existing therapeutic strategies targeting inflammation and metabolic disturbances can promote aging can be broadly grouped into four categories: an overly restrictive epigenetic landscape capable of systemic (blood) factors (including parabiosis), blocking normal differentiation processes or preventing metabolic manipulation (diet regimens [DR] and dietary the induction of tumor suppression programs. restriction mimetics [DRMs]), suppression of cellular Conversely, these cues can induce overly plastic senescence (senolytic drugs), and cellular (partial/ epigenetic states capable of stochastically promoting transient) reprogramming (reviewed in [8]). A inadequate cell fate transitions or activating oncogenic comparison of their underlying mechanisms of action programs [28-31]. This may allow for tissue-specific (e.g., inflammation, nutrient-sensing pathways, epi- responses to changing environmental conditions and, genomic remodeling, autophagy, mitochondria) and perhaps more importantly, it implies that systemic target cells (e.g., stem cells, connective tissue and signals can be interpreted by local injury-repair vasculature cells, senescent cells) immediately suggests mechanisms. For example, whereas defective insulin their potential combinatorial value. Accordingly, it is signaling at the systemic level associates with aging, tempting to suggest that shared mechanisms or target tissue dysfunction, and impaired wound healing, a local cells could be exploited to induce more direct anti-aging potentiation of insulin sensitivity in the liver suffices to www.aging-us.com 6592 AGING switch the hepatic injury response from a stromal repair stem cell niche, to locally integrate systemic signals and process to a more beneficial epithelial repair process alter the type and intensity of repair mechanisms [32]. The ability of the microenvironment, including the strongly suggests that anti-aging strategies should Figure 1. Combinatorial anti-aging strategies: Good and bad directions. (A) Chronic inflammation (or inflamm-aging) is commonly viewed as a central feature in aging and aging-related diseases [6,7,43]. Accordingly, the benefits arising from the anti- inflammatory effects of either dietary interventions/DRM strategies or senolytics might be directly enhanced in combination with well- characterized anti-inflammatory drugs (e.g., aspirin, non-steroidal anti-inflammatory drugs [NSAIDs] or nordihydroguaiaretic acid) [44- 49], which may even operate at the hypothalamus level to impede systemic inflammation-driven aging [50]. However, a new paradigm for the role of inflamm-aging and immunosenescence (an aging-related decline in immune parameters that often leads to subclinical accumulation of pro-inflammatory factors and inflamm-aging) in the aging process begins to suggest that successful aging and longevity can only occur when changes in inflamm-aging are balanced by compensatory anti-inflamm-aging mechanisms [51]. In such a scenario, in which the immune/inflammatory system is adapted/remodeled to provide the best possible anti-pathogen protection when the adaptive immune systems fails in the elderly/aged organisms, the preservation of such apparently detrimental changes may be needed for optimal healthspan/longevity. Therefore, the aforementioned combinations (generating exacerbated decreases of the two components of the immunosenescence/inflamm-aging duo) might cause potential harm in terms of immune response impairment to infections in aged individuals. (B) Because many of the aging-associated features that are reverted by partial cell reprogramming are related to senescence [25], and given that cellular reprogramming on its own has been shown to rejuvenate senescent cells [52], a combination of senolytic and reprogramming strategies might provide synergistic anti-aging effects. However, it should be noted that the presence of inflammatory factors such as interleukin-6 (IL-6) in response to injury-induced senescence promotes cellular plasticity and responsiveness of neighboring cells to in vivo reprogramming-like phenomena [23,24]. Indeed, specific pharmacological and genetic removal of senescent cells has been shown to impair in vivo reprogramming efficiency [53]. The routes, kinetics, and intensities that would distinguish between the beneficial and the harmful effects of senescence in terms of the inherent susceptibility of different cell and tissue types to cellular reprogramming remains an important direction for future studies. (C) The cellular epigenome landscape directs cell fate and reflects its health and biological age. Certain dietary interventions (e.g., ketogenic diets), by altering the availability of key regulators for chromatin- modifying enzymes (e.g., the histone deacetylases inhibitor β-hydroxybutyrate), may provide a direct link between metabolism and epigenomic remodeling to extend healthspan and longevity [54,55]. Given the ability of dietary interventions and DRMs to affect the epigenome [31,56-58] and to suppress the development of senescence [59-62], an intriguing possibility is that dietary, pharmacological, and behavioral strategies targeting nutrient-sensing pathways might synergistically interact with cellular reprogramming strategies to provide better anti-aging outcomes. Nonetheless, certain dietary interventions could promote enhanced stemness and tumorigenicity in specific stem cell compartments and differentiated cell types [63,64]. Interference with normal wound healing processes and potentiation of tumorigenesis and cancer progression might therefore occur after excessive perturbation of the epigenome plasticity, as a maladaptive response to a combination of metabolic manipulations and partial reprogramming. www.aging-us.com 6593 AGING restore a homeostatic balance of sub-cellular trans- the short-range/paracrine communication between ducers (e.g., epigenome, mitochondria), cellular distinct cell types within a tissue, as well as between functions (e.g., stem cells, senescent cells, vascular and different individuals. Unfortunately, we are lacking connective tissue cells –the latter present throughout the robust and standardized approaches to capture such entire organism), and systemic features (e.g., inflam- fundamental stochastic aspects of aging biology. mation, nutrient sensing), in order to dynamically maintain organismal function. Biomathematics to capture the stochasticity of aging biology Endogenous heterogeneity and individual variability might operate as the fundamental epigenetic dimensions Biomathematical and computational systems biology determining the tissue and organismal potential to are expected to deconstruct, model, and simulate the benefit from anti-aging strategies (Figure 2). The power that epigenetic landscapes have on the benefit/risk ratio of anti-aging strategies targeting susceptibility of cells and tissues to retain or lose their inflammation (e.g., senolytics) or nutrient/metabolism- normal identity and function. We recently presented a sensing pathways (e.g., DR/DRMs modulating the theoretical framework in which the acquisition of communication between metabolism and epigenetics) phenotypic plasticity (via transient disruption of the might be dictated by the pre-existing degree of homeostatic resilience of the chromatin structure and epigenetic plasticity and phenotypic malleability of the functioning) followed by reparative differentiation different cell populations in aging tissues. Accordingly, phenomena (replacement or dilution of damaged/ the occurrence of intra- and inter-individual variability diseased cells) might constitute a common route driven by sub-cellular/cellular (e.g., local availability of through which aging-targeted therapies could enhance epigenetic substrates/cofactors and/or metabolic cues) the organismal self-repair capacity after damage, stress, and more systemic features (e.g., inflammation driven and disease [27]. This conceptualization of aging as a by senescent cells, activated fibroblasts or endothelial senescence-inflammation regulatory phenomenon of cells) begins to be understood as a consequence of phenotypic plasticity and reparative reprogramming different aging trajectories occurring in different tissue actually incorporates the concept of frailty, which can cell subpopulations [33]. This likely reflects changes in be viewed not only as a deficit accumulation, but also Figure 2. Biomathematical approaches to accelerate the performance and personalize the use of aging-targeted strategies. The benefit/risk ratio of aging-targeted interventions largely depends on intra- and inter-individual variations of susceptibility to the healthspan-, resilience-, and/or lifespan-promoting effects of the interventions. Stochastic biomathematical modeling and computational simulation platforms should be developed as in silico strategies to accelerate the performance of clinical trials targeting human aging, and to provide personalized approaches and robust biomarkers of healthy aging at the individual-to-population levels. www.aging-us.com 6594 AGING as a reduction in the molecular/cellular/physiological ment, tissue-specific fibrosis or dysfunction (including reserves of the aged organs/organism, leading to less tumorigenesis) because of loss of cellular identity. Such efficient responses to stress. A predictive computational stochastic biomathematical platforms would comprise a model of cell fate reprogramming revealed that machine-learning/computational component to uncover pathological cell states, and their suppression, might the phenotypic behavior of cell populations mimicking result from the sole stochastic translation of metabolic the intra- and inter-individual heterogeneity and cues (e.g., DR and DRMs) into resilient/plastic cell variability when challenged with aging-targeting states [34]. A more advanced stochastic, multiscale strategies (Figure 2). Computational tools suited for reduction method of combined epigenetic regulation- complex stochastic dynamical systems could be gene regulatory network has mathematically captured employed to in silico construct and comparatively how epigenetic heterogeneity can operate as the driving integrate the shape and evolution of epigenome force governing the routes and kinetics to entry and exit remodeling upon such imposition of aging-targeted from unrestrained epigenetic plastic states [35]. approaches. Quantitative probabilistic simulations could Moreover, we were able to computationally validate the be then incorporated to functionally infer the propor- likelihood of unlocking aberrant cell states disabled for tions and interconversion rates of the existing com- reparative differentiation and prone to malignant munities of epi-states. Ideally, a multi-scale theoretical transformation (and drive their function to their correct integration of epi-phenotype dynamics prospectively repair function) solely by manipulating the intensity and coupled to confirmatory laboratory-based testing of the direction of a few epigenetic control switches [35]. in vivo amelioration of aging hallmarks, would allow an Altogether, these findings provided support to the iterative improvement process capable not only of notion that predictive mathematical modeling and predicting the effectiveness but also of personalizing computational simulation platforms might be aging-targeting interventions. Indeed, by computa- incorporated as new tools to optimize the clinical tionally determining the key regulatory nodes and success of aging-targeting therapeutics while providing kinetic routes of phenotypic plasticity in an unbiased individualized ways to target human aging. By and systematic manner, stochastic biomathematical analyzing computationally-generated proxies of the platforms would provide end-points and biomarkers of impact of aging-targeted strategies on the behavior of healthy aging when assessing the individual-to- epigenetic-regulatory (ER) systems controlling cell fate population levels of response to aging-targeted inter- reprogramming, we here exemplify how baseline ventions. epigenetic heterogeneity might dictate not only the efficacy of single anti-aging therapies such as ketogenic The unique features of clinical trials targeting diet (as an archetypal lifespan/healthspan-promoting human aging dietary approach leading to inhibition of histone deacetylases [36-38]) and mTOR inhibition (as an The ultimate goal of aging-targeted strategies extends archetypal lifespan/healthspan-promoting metabolic beyond an isolated impact in each separate age-related approach that couples nutrient sensing to epigenetic disease. Measuring the global impact of such regulation by promoting a decrease in histone healthspan-/resilience-/lifespan-promoting interventions acetylation [39,40]) but also the degree of an in silico on “hard” composite outcomes such as burden of predicted anti-aging synergistic response to a ketogenic- chronic diseases, functional dependence, and/or like dietary regimen concurrent with mTOR blockade mortality, will require a major effort in the design of (Figure 3). controlled clinical trials in large, heterogeneous, elderly populations. Before committing to complex, long-term Stochastic biomathematical platforms to strengthen studies, a series of intermediate, smaller clinical trials of the efficacy and personalization of healthspan/ a given aging-targeted intervention should provide resilience/lifespan-promoting therapies experimental evidence for the concept that basic aging- driving mechanisms can be therapeutically targeted. The aforementioned computational models are Moreover, because aging is not an indication currently consistent with a scenario in which cell/tissue plasticity recognized by the reference regulatory agencies (FDA, and resilience are fundamental epigenetic phenomena EMA), a key goal for forthcoming clinical trials will be associated with phenotypic variation that drives to define outcomes representative of fundamental aging successful repair, degeneration or transformation. features that might be acceptable as potential Because such models predict that the fine-tuning of indications. Such in human amelioration of aging certain signals might re-direct plastic epi-states into hallmarks might include multi-morbidity, frailty and phenotypic resilience or vulnerability, they seem to be functional decline. In this regard, two broad strategies idoneous to in silico test the ability of a given aging- involving short-duration or pilot studies have been targeted intervention to promote tissue repair impair- proposed to solidify the rationale and better inform the www.aging-us.com 6595 AGING Figure 3. Baseline epigenetic heterogeneity and efficacy of anti-aging strategies: A proof-of-concept in silico trial. We have recently presented a stochastic biomathematical modeling and computational simulation strategy that might be incorporated as a valuable tool for assessing the benefit/risk ratio of therapeutic approaches aimed to target the aging/cancer-related perturbations of the epigenome [35]. Briefly, we quantified the heterogeneity and robustness of differentiation epigenetic and pluripotency regulatory systems (DERS and PERS, respectively) in terms of the average reprogramming time associated with differentiation-primed (benefit) and pluripotency-locked (risk) states. Such calculation can therefore be employed as a proxy of the expected efficacy of a given anti-aging strategy (i.e., longer reprogramming times associate with more efficient anti-aging outcomes). Regarding DERS heterogeneity, we observed three different clusters associated with a differentiated-primed behavior (DERS1), a differentiation-refractory (stem-like) behavior (DERS2), and an indecision behavior (DERS3). Several kinetic parameters associated with histone deacetylase (HDAC, c11) and histone acetylase (HAC, c15 and c16) activities sufficed to account for the distinction between DERS clusters [35]. Regarding PERS heterogeneity, we observed three different clusters defined by the large (PERS1), intermediate (PERS2), and small (PERS3) values of the average waiting time needed for transitioning from closed to open ER states [35]. As a proof of concept, here we considered nine combinatorial scenarios corresponding to all the possible combinations between three different differentiation DERS and PERS challenged with three different anti-aging strategies (A), namely ketogenic diet (mimicked by reducing by 50% the kinetic parameter associated with HDAC activity [i.e., c11]), NAD+ boosters (mimicked by increasing by 50% the kinetic parameter associated with HDAC activity [i.e., c11]), and mTOR inhibition (mimicked by reducing by 25% the two parameters associated with HAC activity [i.e., c15 and c16]). Whereas ketogenic diet- and mTOR inhibition-like strategies were notably found to be more efficient than NAD+ boosting-like approaches across all the baseline epigenetic scenarios, it was noteworthy that the most sensitive one was that defined by the combination of PERS3 with DERS1 (A). Based on these results, we decided to assess the direction and intensity of the average reprogramming time when combining ketogenic diet- and mTOR inhibition-like strategies (B). With the exception of baseline scenarios where the strong single-agent positive outcome left little room for additional gains, we observed that strong, synergistic interactions tend to be more specific to particular epigenetic states than were single strategies (B). Our in silico approach exemplifies how baseline epigenetic heterogeneity might dictate not only the positive outcomes of single aging-targeted therapies such as ketogenic diet and mTOR inhibition but also the degree of positive synergistic effects that were predicted to occur upon concurrent targeting of an apparently existing HDAC-mTOR cross-link. www.aging-us.com 6596 AGING design and scale of larger and longer trials [2]. The so- Aging-targeted interventions: challenging an old called “extending healthspan” study design is based on paradigm the notion that, if aging is associated with a gradual accumulation of multiple chronic diseases, geriatric The conventional paradigm “one disease, one drug” syndromes, and functional decline, those interventions should be updated to achieve the vision of targeting directly targeting the aging machinery should have a aging as a common component of human diseases. The significant effect on multi-morbidity. Conversely, the current deterministic genetic paradigm of diagnosing so-called “enhancing resilience” study design is based and treating each separate age-related disease fails to fit on the notion that, if the ability of response to or with the broader anti-aging strategies aimed to address recovery from an acute health stress back to the the closely related concepts of healthspan, resilience, functional/healthy baseline declines with age, those and lifespan, which should be therapeutically managed interventions directly targeting the aging machinery in the absence of discrete, targetable genetic drivers of should be able to enhance the capacity of an individual to aging progression. Perhaps more importantly, current recover functional independence. The question remains frameworks cannot capture the stochastic aspects that as to whether rejuvenation interventions contributing to drive the shared trade-offs of the emerging strategies for “lifespan extension” such as transient reprogramming, organismal healthspan and rejuvenation, namely tissue- which has been observed exclusively in the context of repair/wound-healing impairment and tumorigenesis. premature aging [25, 26], may also prove beneficial for whole-organism rejuvenation in the absence of injury or Successful clinical trials with new families of candidate disease and/or in a naturally aged context. interventions targeting the biologic machinery of aging per se would be groundbreaking; delaying, preventing Even in the context of broad strategies such as the (or even reversing) the aging process would result in proof-of-concept studies described above, the sole tremendous cost savings for healthcare systems while testing of existing drug pipelines (e.g., sirtuin agonists, increasing the productive contributions that could be senolytic agents, DRMs including metformin, made by the older members of our societies. By mitochondrial function and autophagy regulators, anti- modeling and predicting the behavior of interventions hypertensive agents, etc) and non-pharmacological that target the aging hallmarks in both long-term and strategies (e.g., blood factors, DRs, partial cellular acute settings, defined by extension of healthspan/ reprograming, inducible telomerase, exercise, fecal lifespan and enhanced resilience to acute stressors (i.e., microbiota transplantation, etc) might generate reduced frailty), respectively, robust and standardized significant challenges for clinical trial design regarding approaches such as stochastic biomathematical population selection, study length and type of platforms would have the ability to sidestep most of the intervention, and measurement of the various outcomes current challenges in aging-targeting clinical trials, to for each candidate. Accordingly, a recent study has accelerate the achievement of optimum health and life raised concerns about broad recommendations for the quality in aging populations. use of metformin as a potential healthspan extending treatment [10,41] in healthy individuals [42]. The ACKNOWLEDGEMENTS influence of metformin on aerobic exercise training- induced improvements in physiological functions in The authors would like to thank Dr. Kenneth McCreath older adults was highly variable, with the occurrence of for editorial support. positive and negative responders being associated with the heterogenous effect of metformin on the mito- CONFLICTS OF INTEREST chondria [42]. These data strongly support the notion that prior to prescribing any aging-targeted intervention, The authors have no conflict of interests to declare. additional studies are needed to understand the mechanisms that elicit positive and negative responses FUNDING before assuming synergism of combining such strategies. In this scenario, we propose that the Work in the Menendez laboratory is supported by the development of predictive mathematical models and Spanish Ministry of Science and Innovation (Grant computational simulation platforms to operatively SAF2016-80639-P, Plan Nacional de l+D+I, founded by integrate the multi-scale epi-phenotypes from single the European Regional Development Fund, Spain) and cells to individuals and populations might enhance the by an unrestricted research grant from the Fundació clinical performance of aging-targeting therapeutics by Oncolliga Girona (Lliga catalana d’ajuda al malalt de helping to circumvent the trade-offs of aging-targeted càncer, Girona). This work is supported also by a grant interventions while providing individualized ways to of the Obra Social La Caixa Foundation on target human aging (Figure 3). Collaborative Mathematics awarded to the Centre de www.aging-us.com 6597 AGING Recerca Matemàtica (CRM). The authors have been Geroscience: linking aging to chronic disease. Cell. partially funded by the CERCA Programme of the 2014; 159:709–13. Generalitat de Catalunya. E.C. is the recipient of a Sara https://doi.org/10.1016/j.cell.2014.10.039 Borrell post-doctoral contract (CD15/00033, Ministerio PMID:25417146 de Sanidad y Consumo, Fondo de Investigación 8. Mahmoudi S, Xu L, Brunet A. Turning back time with Sanitaria, Spain). N.F-B. and T.A. acknowledge emerging rejuvenation strategies. Nat Cell Biol. 2019; MINECO and AGAUR for funding under Grants 21:32–43. https://doi.org/10.1038/s41556-018-0206-0 MTM2015-71509-C2-1-R and 2014SGR1307. T.A. PMID:30602763 acknowledges support from MINECO for funding awarded to the Barcelona Graduate School of 9. López-Otín C, Galluzzi L, Freije JM, Madeo F, Kroemer Mathematics under the “Marı́a de Maeztu” programme G. Metabolic Control of Longevity. Cell. 2016; (MDM-2014-0445). 166:802–21. https://doi.org/10.1016/j.cell.2016.07.031 REFERENCES PMID:27518560 10. Barzilai N, Crandall JP, Kritchevsky SB, Espeland MA. 1. Brown GC. Living too long: the current focus of Metformin as a Tool to Target Aging. Cell Metab. medical research on increasing the quantity, rather 2016; 23:1060–65. than the quality, of life is damaging our health and https://doi.org/10.1016/j.cmet.2016.05.011 harming the economy. EMBO Rep. 2015; 16:137–41. PMID:27304507 https://doi.org/10.15252/embr.201439518 PMID:25525070 11. Lamming DW, Ye L, Sabatini DM, Baur JA. Rapalogs 2. Newman JC, Milman S, Hashmi SK, Austad SN, and mTOR inhibitors as anti-aging therapeutics. J Clin Kirkland JL, Halter JB, Barzilai N. Strategies and Invest. 2013; 123:980–89. Challenges in Clinical Trials Targeting Human Aging. J https://doi.org/10.1172/JCI64099 PMID:23454761 Gerontol A Biol Sci Med Sci. 2016; 71:1424–34. 12. Demaria M, Ohtani N, Youssef SA, Rodier F, Toussaint https://doi.org/10.1093/gerona/glw149 W, Mitchell JR, Laberge RM, Vijg J, Van Steeg H, Dollé PMID:27535968 ME, Hoeijmakers JH, de Bruin A, Hara E, Campisi J. An 3. Espeland MA, Crimmins EM, Grossardt BR, Crandall essential role for senescent cells in optimal wound JP, Gelfond JA, Harris TB, Kritchevsky SB, Manson JE, healing through secretion of PDGF-AA. Dev Cell. 2014; Robinson JG, Rocca WA, Temprosa M, Thomas F, 31:722–33. Wallace R, Barzilai N, and Multimorbidity Clinical https://doi.org/10.1016/j.devcel.2014.11.012 Trials Consortium. Clinical Trials Targeting Aging and PMID:25499914 Age-Related Multimorbidity. J Gerontol A Biol Sci 13. Calcinotto A, Kohli J, Zagato E, Pellegrini L, Demaria Med Sci. 2017; 72:355–61. M, Alimonti A. Cellular Senescence: Aging, Cancer, https://doi.org/10.1093/gerona/glw220 and Injury. Physiol Rev. 2019; 99:1047–78. PMID:28364543 https://doi.org/10.1152/physrev.00020.2018 4. Barzilai N, Cuervo AM, Austad S. Aging as a Biological PMID:30648461 Target for Prevention and Therapy. JAMA. 2018; 14. Krizhanovsky V, Yon M, Dickins RA, Hearn S, Simon J, 320:1321–22. Miething C, Yee H, Zender L, Lowe SW. Senescence of https://doi.org/10.1001/jama.2018.9562 activated stellate cells limits liver fibrosis. Cell. 2008; PMID:30242337 134:657–67. 5. Mahmoudi S, Brunet A. Aging and reprogramming: a https://doi.org/10.1016/j.cell.2008.06.049 two-way street. Curr Opin Cell Biol. 2012; 24:744–56. PMID:18724938 https://doi.org/10.1016/j.ceb.2012.10.004 15. Ovadya Y, Krizhanovsky V. Strategies targeting cellular PMID:23146768 senescence. J Clin Invest. 2018; 128:1247–54. 6. López-Otín C, Blasco MA, Partridge L, Serrano M, https://doi.org/10.1172/JCI95149 PMID:29608140 Kroemer G. The hallmarks of aging. Cell. 2013; 16. Schafer MJ, White TA, Iijima K, Haak AJ, Ligresti G, 153:1194–217. Atkinson EJ, Oberg AL, Birch J, Salmonowicz H, Zhu Y, https://doi.org/10.1016/j.cell.2013.05.039 Mazula DL, Brooks RW, Fuhrmann-Stroissnigg H, et al. PMID:23746838 Cellular senescence mediates fibrotic pulmonary 7. Kennedy BK, Berger SL, Brunet A, Campisi J, Cuervo disease. Nat Commun. 2017; 8:14532. AM, Epel ES, Franceschi C, Lithgow GJ, Morimoto RI, https://doi.org/10.1038/ncomms14532 Pessin JE, Rando TA, Richardson A, Schadt EE, et al. PMID:28230051 www.aging-us.com 6598 AGING 17. Wang R, Yu Z, Sunchu B, Shoaf J, Dang I, Zhao S, 1733.e12. https://doi.org/10.1016/j.cell.2016.11.052 Caples K, Bradley L, Beaver LM, Ho E, Löhr CV, Perez PMID:27984723 VI. Rapamycin inhibits the secretory phenotype of 26. Ocampo A, Reddy P, Belmonte JC. Anti-Aging senescent cells by a Nrf2-independent mechanism. Strategies Based on Cellular Reprogramming. Trends Aging Cell. 2017; 16:564–74. Mol Med. 2016; 22:725–38. https://doi.org/10.1111/acel.12587 PMID:28371119 https://doi.org/10.1016/j.molmed.2016.06.005 18. Demidenko ZN, Zubova SG, Bukreeva EI, Pospelov VA, PMID:27426043 Pospelova TV, Blagosklonny MV. Rapamycin 27. Menendez JA, Alarcón T. Senescence-Inflammatory decelerates cellular senescence. Cell Cycle. 2009; Regulation of Reparative Cellular Reprogramming in 8:1888–95. https://doi.org/10.4161/cc.8.12.8606 Aging and Cancer. Front Cell Dev Biol. 2017; 5:49. PMID:19471117 https://doi.org/10.3389/fcell.2017.00049 19. Campisi J. Aging, cellular senescence, and cancer. PMID:28529938 Annu Rev Physiol. 2013; 75:685–705. 28. Menendez JA, Alarcón T. Metabostemness: a new https://doi.org/10.1146/annurev-physiol-030212- cancer hallmark. Front Oncol. 2014; 4:262. 183653 PMID:23140366 https://doi.org/10.3389/fonc.2014.00262 20. Muñoz-Espín D, Serrano M. Cellular senescence: from PMID:25325014 physiology to pathology. Nat Rev Mol Cell Biol. 2014; 29. Pal S, Tyler JK. Epigenetics and aging. Sci Adv. 2016; 15:482–96. https://doi.org/10.1038/nrm3823 2:e1600584. https://doi.org/10.1126/sciadv.1600584 PMID:24954210 PMID:27482540 21. Ohnishi K, Semi K, Yamamoto T, Shimizu M, Tanaka A, 30. Flavahan WA, Gaskell E, Bernstein BE. Epigenetic Mitsunaga K, Okita K, Osafune K, Arioka Y, Maeda T, plasticity and the hallmarks of cancer. Science. 2017; Soejima H, Moriwaki H, Yamanaka S, et al. Premature 357:eaal2380. termination of reprogramming in vivo leads to cancer https://doi.org/10.1126/science.aal2380 development through altered epigenetic regulation. PMID:28729483 Cell. 2014; 156:663–77. https://doi.org/10.1016/j.cell.2014.01.005 31. Reid MA, Dai Z, Locasale JW. The impact of cellular PMID:24529372 metabolism on chromatin dynamics and epigenetics. Nat Cell Biol. 2017; 19:1298–306. 22. Abad M, Mosteiro L, Pantoja C, Cañamero M, Rayon https://doi.org/10.1038/ncb3629 PMID:29058720 T, Ors I, Graña O, Megías D, Domínguez O, Martínez D, Manzanares M, Ortega S, Serrano M. 32. Manzano-Núñez F, Arámbul-Anthony MJ, Galán Reprogramming in vivo produces teratomas and iPS Albiñana A, Leal Tassias A, Acosta Umanzor C, cells with totipotency features. Nature. 2013; Borreda Gascó I, Herrera A, Forteza Vila J, Burks DJ, 502:340–45. https://doi.org/10.1038/nature12586 Noon LA. Insulin resistance disrupts epithelial repair PMID:24025773 and niche-progenitor Fgf signaling during chronic liver injury. PLoS Biol. 2019; 17:e2006972. 23. Mosteiro L, Pantoja C, Alcazar N, Marión RM, https://doi.org/10.1371/journal.pbio.2006972 Chondronasiou D, Rovira M, Fernandez-Marcos PJ, PMID:30695023 Muñoz-Martin M, Blanco-Aparicio C, Pastor J, Gómez- López G, De Martino A, Blasco MA, et al. Tissue 33. Mahmoudi S, Mancini E, Moore A, Xu L, Jahanbani F, damage and senescence provide critical signals for Hebestreit K, Srinivasan R, Li X, Devarajan K, Prelot L, cellular reprogramming in vivo. Science. 2016; Ang CE, Shibuya Y, Benayoun BA, et al. Old fibroblasts 354:aaf4445. secrete inflammatory cytokines that drive variability https://doi.org/10.1126/science.aaf4445 in reprogramming efficiency and may affect wound PMID:27884981 healing between old individuals. bioRxiv. 2018; https://doi.org/10.1101/448431 24. Mosteiro L, Pantoja C, de Martino A, Serrano M. Senescence promotes in vivo reprogramming through 34. Folguera-Blasco N, Cuyàs E, Menéndez JA, Alarcón T. p16INK4a and IL-6. Aging Cell. 2018; 17:e12711. Epigenetic regulation of cell fate reprogramming in https://doi.org/10.1111/acel.12711 PMID:29280266 aging and disease: A predictive computational model. PLOS Comput Biol. 2018; 14:e1006052. 25. Ocampo A, Reddy P, Martinez-Redondo P, Platero- https://doi.org/10.1371/journal.pcbi.1006052 Luengo A, Hatanaka F, Hishida T, Li M, Lam D, Kurita PMID:29543808 M, Beyret E, Araoka T, Vazquez-Ferrer E, Donoso D, et al. In Vivo Amelioration of Age-Associated Hallmarks 35. Folguera-Blasco N, Pérez-Carrasco R, Cuyàs E, by Partial Reprogramming. Cell. 2016; 167:1719– Menendez JA, Alarcón T. A multiscale model of www.aging-us.com 6599 AGING epigenetic heterogeneity-driven cell fate decision- N Y Acad Sci. 2000; 908:244–54. making. PLOS Comput Biol. 2019; 15:e1006592. https://doi.org/10.1111/j.1749-6632.2000.tb06651.x https://doi.org/10.1371/journal.pcbi.1006592 PMID:10911963 PMID:31039148 44. Youm YH, Grant RW, McCabe LR, Albarado DC, 36. Rajendran P, Williams DE, Ho E, Dashwood RH. Nguyen KY, Ravussin A, Pistell P, Newman S, Carter R, Metabolism as a key to histone deacetylase Laque A, Münzberg H, Rosen CJ, Ingram DK, et al. inhibition. Crit Rev Biochem Mol Biol. 2011; 46:181– Canonical Nlrp3 inflammasome links systemic low- 99. https://doi.org/10.3109/10409238.2011.557713 grade inflammation to functional decline in aging. Cell PMID:21599534 Metab. 2013; 18:519–32. https://doi.org/10.1016/j.cmet.2013.09.010 37. Newman JC, Verdin E. Ketone bodies as signaling PMID:24093676 metabolites. Trends Endocrinol Metab. 2014; 25:42– 52. https://doi.org/10.1016/j.tem.2013.09.002 45. Youm YH, Nguyen KY, Grant RW, Goldberg EL, PMID:24140022 Bodogai M, Kim D, D’Agostino D, Planavsky N, Lupfer C, Kanneganti TD, Kang S, Horvath TL, Fahmy TM, et 38. Roberts MN, Wallace MA, Tomilov AA, Zhou Z, al. The ketone metabolite β-hydroxybutyrate blocks Marcotte GR, Tran D, Perez G, Gutierrez-Casado E, NLRP3 inflammasome-mediated inflammatory Koike S, Knotts TA, Imai DM, Griffey SM, Kim K, et al. A Ketogenic Diet Extends Longevity and Healthspan in disease. Nat Med. 2015; 21:263–69. https://doi.org/10.1038/nm.3804 PMID:25686106 Adult Mice. Cell Metab. 2017; 26:539–546.e5. https://doi.org/10.1016/j.cmet.2017.08.005 46. Strong R, Miller RA, Astle CM, Floyd RA, Flurkey K, PMID:28877457 Hensley KL, Javors MA, Leeuwenburgh C, Nelson JF, Ongini E, Nadon NL, Warner HR, Harrison DE. 39. Makarević J, Tawanaie N, Juengel E, Reiter M, Mani J, Nordihydroguaiaretic acid and aspirin increase Tsaur I, Bartsch G, Haferkamp A, Blaheta RA. Cross- communication between histone H3 and H4 lifespan of genetically heterogeneous male mice. Aging Cell. 2008; 7:641–50. acetylation and Akt-mTOR signalling in prostate https://doi.org/10.1111/j.1474-9726.2008.00414.x cancer cells. J Cell Mol Med. 2014; 18:1460–66. PMID:18631321 https://doi.org/10.1111/jcmm.12299 PMID:24779401 47. Hundal RS, Petersen KF, Mayerson AB, Randhawa PS, 40. Workman JJ, Chen H, Laribee RN. Saccharomyces Inzucchi S, Shoelson SE, Shulman GI. Mechanism by cerevisiae TORC1 Controls Histone Acetylation by Signaling Through the Sit4/PP6 Phosphatase to which high-dose aspirin improves glucose metabolism in type 2 diabetes. J Clin Invest. 2002; 109:1321–26. Regulate Sirtuin Deacetylase Nuclear Accumulation. https://doi.org/10.1172/JCI0214955 PMID:12021247 Genetics. 2016; 203:1733–46. https://doi.org/10.1534/genetics.116.188458 48. Gasparini L, Ongini E, Wenk G. Non-steroidal anti- PMID:27343235 inflammatory drugs (NSAIDs) in Alzheimer’s disease: old and new mechanisms of action. J Neurochem. 41. Cuyàs E, Fernández-Arroyo S, Buxó M, Pernas S, 2004; 91:521–36. https://doi.org/10.1111/j.1471- Dorca J, Álvarez I, Martínez S, Pérez-Garcia JM, 4159.2004.02743.x PMID:15485484 Batista-López N, Rodríguez-Sánchez CA, Amillano K, Domínguez S, Luque M, et al. Metformin induces a 49. Wan QL, Zheng SQ, Wu GS, Luo HR. Aspirin extends fasting- and antifolate-mimicking modification of the lifespan of Caenorhabditis elegans via AMPK and systemic host metabolism in breast cancer patients. DAF-16/FOXO in dietary restriction pathway. Exp Aging (Albany NY). 2019; 11:2874–88. Gerontol. 2013; 48:499–506. https://doi.org/10.18632/aging.101960 https://doi.org/10.1016/j.exger.2013.02.020 PMID:31076561 PMID:23485446 42. Konopka AR, Laurin JL, Schoenberg HM, Reid JJ, 50. Zhang G, Li J, Purkayastha S, Tang Y, Zhang H, Yin Y, Li Castor WM, Wolff CA, Musci RV, Safairad OD, Linden B, Liu G, Cai D. Hypothalamic programming of MA, Biela LM, Bailey SM, Hamilton KL, Miller BF. systemic ageing involving IKK-β, NF-κB and GnRH. Metformin inhibits mitochondrial adaptations to Nature. 2013; 497:211–16. aerobic exercise training in older adults. Aging Cell. https://doi.org/10.1038/nature12143 2019; 18:e12880. https://doi.org/10.1111/acel.12880 PMID:23636330 PMID:30548390 51. Fulop T, Larbi A, Dupuis G, Le Page A, Frost EH, Cohen 43. Franceschi C, Bonafè M, Valensin S, Olivieri F, De Luca AA, Witkowski JM, Franceschi C. Immunosenescence M, Ottaviani E, De Benedictis G. Inflamm-aging. An and Inflamm-Aging As Two Sides of the Same Coin: evolutionary perspective on immunosenescence. Ann friends or Foes? Front Immunol. 2018; 8:1960. www.aging-us.com 6600 AGING https://doi.org/10.3389/fimmu.2017.01960 60. Laberge RM, Sun Y, Orjalo AV, Patil CK, Freund A, PMID:29375577 Zhou L, Curran SC, Davalos AR, Wilson-Edell KA, Liu S, 52. Lapasset L, Milhavet O, Prieur A, Besnard E, Babled A, Limbad C, Demaria M, Li P, et al. MTOR regulates the pro-tumorigenic senescence-associated secretory Aït-Hamou N, Leschik J, Pellestor F, Ramirez JM, De phenotype by promoting IL1A translation. Nat Cell Vos J, Lehmann S, Lemaitre JM. Rejuvenating Biol. 2015; 17:1049–61. senescent and centenarian human cells by https://doi.org/10.1038/ncb3195 PMID:26147250 reprogramming through the pluripotent state. Genes 61. Herranz N, Gallage S, Mellone M, Wuestefeld T, Klotz Dev. 2011; 25:2248–53. S, Hanley CJ, Raguz S, Acosta JC, Innes AJ, Banito A, https://doi.org/10.1101/gad.173922.111 PMID:22056670 Georgilis A, Montoya A, Wolter K, et al. mTOR regulates MAPKAPK2 translation to control the 53. Chiche A, Le Roux I, von Joest M, Sakai H, Aguín SB, senescence-associated secretory phenotype. Nat Cell Cazin C, Salam R, Fiette L, Alegria O, Flamant P, Biol. 2015; 17:1205–17. Tajbakhsh S, Li H. Injury-Induced Senescence Enables https://doi.org/10.1038/ncb3225 PMID:26280535 In Vivo Reprogramming in Skeletal Muscle. Cell Stem 62. Han X, Tai H, Wang X, Wang Z, Zhou J, Wei X, Ding Y, Cell. 2017; 20:407–414.e4. Gong H, Mo C, Zhang J, Qin J, Ma Y, Huang N, et al. https://doi.org/10.1016/j.stem.2016.11.020 PMID:28017795 AMPK activation protects cells from oxidative stress- induced senescence via autophagic flux restoration 54. Newman JC, Covarrubias AJ, Zhao M, Yu X, Gut P, Ng and intracellular NAD(+) elevation. Aging Cell. 2016; CP, Huang Y, Haldar S, Verdin E. Ketogenic Diet 15:416–27. https://doi.org/10.1111/acel.12446 Reduces Midlife Mortality and Improves Memory in PMID:26890602 Aging Mice. Cell Metab. 2017; 26:547–557.e8. 63. Beyaz S, Mana MD, Roper J, Kedrin D, Saadatpour A, https://doi.org/10.1016/j.cmet.2017.08.004 PMID:28877458 Hong SJ, Bauer-Rowe KE, Xifaras ME, Akkad A, Arias E, Pinello L, Katz Y, Shinagare S, et al. High-fat diet 55. Blagosklonny MV. The mystery of the ketogenic diet: enhances stemness and tumorigenicity of intestinal benevolent pseudo-diabetes. Cell Cycle. 2019; progenitors. Nature. 2016; 531:53–58. 18:2157-63. https://doi.org/10.1038/nature17173 https://doi.org/10.1080/15384101.2019.1644765 PMID:26935695 PMID: 31368400 64. Schwitalla S, Fingerle AA, Cammareri P, Nebelsiek T, 56. Benayoun BA, Pollina EA, Brunet A. Epigenetic Göktuna SI, Ziegler PK, Canli O, Heijmans J, Huels DJ, regulation of ageing: linking environmental inputs to Moreaux G, Rupec RA, Gerhard M, Schmid R, et al. genomic stability. Nat Rev Mol Cell Biol. 2015; Intestinal tumorigenesis initiated by dedifferentiation 16:593–610. https://doi.org/10.1038/nrm4048 and acquisition of stem-cell-like properties. Cell. PMID:26373265 2013; 152:25–38. 57. Kim CH, Lee EK, Choi YJ, An HJ, Jeong HO, Park D, Kim https://doi.org/10.1016/j.cell.2012.12.012 BC, Yu BP, Bhak J, Chung HY. Short-term calorie PMID:23273993 restriction ameliorates genomewide, age-related alterations in DNA methylation. Aging Cell. 2016; 15:1074–81. https://doi.org/10.1111/acel.12513 PMID:27561685 58. Hahn O, Grönke S, Stubbs TM, Ficz G, Hendrich O, Krueger F, Andrews S, Zhang Q, Wakelam MJ, Beyer A, Reik W, Partridge L. Dietary restriction protects from age-associated DNA methylation and induces epigenetic reprogramming of lipid metabolism. Genome Biol. 2017; 18:56. https://doi.org/10.1186/s13059-017-1187-1 PMID:28351387 59. Tran D, Bergholz J, Zhang H, He H, Wang Y, Zhang Y, Li Q, Kirkland JL, Xiao ZX. Insulin-like growth factor-1 regulates the SIRT1-p53 pathway in cellular senescence. Aging Cell. 2014; 13:669–78. https://doi.org/10.1111/acel.12219 PMID:25070626 www.aging-us.com 6601 AGING