Regulation of senescence traits by MAPKs

Abstract

A phenotype of indefinite growth arrest acquired in response to sublethal damage, cellular senescence affects normal aging and age-related disease. Mitogen-activated protein kinases (MAPKs) are capable of sensing changes in cellular conditions, and in turn elicit adaptive responses including cell senescence. MAPKs modulate the levels and function of many proteins, including proinflammatory factors and factors in the p21/p53 and p16/RB pathways, the main senescence-regulatory axes. Through these actions, MAPKs implement key traits of senescence—growth arrest, cell survival, and the senescence-associated secretory phenotype (SASP). In this review, we summarize and discuss our current knowledge of the impact of MAPKs in senescence. In addition, given that eliminating or suppressing senescent cells can improve health span, we discuss the function and possible exploitation of MAPKs in the elimination (senolysis) or suppression (senostasis) of senescent cells.

Introduction

Cellular senescence is a program implemented by cells responding to a variety of stresses that cause macromolecular damage. In turn, cells that become senescent exhibit long-term growth arrest and the senescence-associated secretory phenotype (SASP), through which cells secrete proinflammatory and tissue-remodeling factors that have local and systemic impacts (Gorgoulis et al. 2019). Senescence has been found to be both beneficial and detrimental for organ homeostasis (He and Sharpless 2017). Among its benefits, senescence contributes to embryonic development, wound healing, and tumor suppression in young persons (Collado and Serrano 2010; Munoz-Espin et al. 2013; Storer et al. 2013; Demaria et al. 2014). On the other hand, the adverse effects of senescent cells accumulating in tissues are often apparent with advancing age, as they exacerbate age-related pathologies including cancer, sarcopenia, diabetes, and Alzheimer’s disease (Campisi 2013; Lopez-Otin et al. 2013; van Deursen 2014; McHugh and Gil 2018). Given the harmful influence of senescent cells during aging, there is much interest in clearing senescent cells therapeutically through genetic and pharmacologic approaches (Demaria et al. 2014; Baker et al. 2016; Chang et al. 2016). While the clinical usefulness of current genetic approaches to intervene in senescence (discussed by Soto-Gamez and Demaria 2017; McHugh and Gil 2018) are limited, there has been an escalation of efforts to identify chemical senolytic and senomorphic/senostatic interventions to combat age-associated diseases (Childs et al. 2014; Zhu et al. 2015; Yosef et al. 2016; Kirkland and Tchkonia 2017).

To develop rational approaches directed at senescent cells, it is essential to understand the molecular programs that enable the various senescence traits. The best-studied senescence-associated signaling pathways are triggered by damage to DNA and other cellular components that activate the p53 (TP53)-p21 (CDKN1A) axis, the p16 (CDKN2A)-retinoblastoma (RB) axis, and the secretion of SASP factors (Lujambio 2016; Soto-Gamez and Demaria 2017; Herranz and Gil 2018). While these pathways are robustly influenced by the PI3K-mTOR (phosphoinositide 3-kinase-mammalian target of rapamycin) signaling kinases, the regulation of senescent traits by MAPK (mitogen-activated protein kinase) cascades is becoming increasingly apparent (Xu et al. 2014; Martinez-Zamudio et al. 2017). The three main classes of MAPKs include ERK1/2 (extracellular signal-regulated kinase), JNK (c-Jun N-terminal kinase), and p38; among these, ERK1/2 and p38 are most closely linked to cellular senescence (Sun et al. 2007; Debacq-Chainiaux et al. 2010; Passos et al. 2010; Freund et al. 2011; Deschenes-Simard et al. 2013; Storer et al. 2013). Although JNK appears to have a less prominent role in senescence, recent reports have implicated this MAPK in senescent traits (Lee et al. 2010; Spallarossa et al. 2010; Yosef et al. 2017; Vizioli et al. 2020). Key aspects of the influence of MAPKs on growth suppression, resistance to apoptosis, and other traits of senescent cells have only emerged in recent years (Fig. 1) (Ziaei et al. 2012; Herranz et al. 2015; Culerrier et al. 2016; Slobodnyuk et al. 2019). Here, we review our knowledge and discuss the role of MAPKs on senescence traits.

Fig. 1.

Known implications of MAPK pathways in cellular senescence traits. After exposure to senescence-inducing stimuli, MAPK pathways (mainly p38 and ERK1/2) drive the implementation of the senescence phenotype. MAPKs exert direct control over the main senescence traits—cell survival, cell cycle arrest, and the senescence-associated secretory phenotype (SASP). The main MAPK regulatory events thus far implicated in senescence are indicated.

MAPK networks in cellular senescence

Discovered in the early 1990s, MAPKs represent major signaling cascades in cell biology (Pearson et al. 2001). This superfamily of proteins is mainly comprised of kinases that mediate chains of phosphorylation events. Using simplified terminology, membrane receptors activated by mitogens, cytokines, and stress agents activate MAPKKKKs, which phosphorylate MAPKKKs, which in turn phosphorylate MAPKKs, and these then phosphorylate MAPKs. Downstream effectors of MAPKs include several proteins such as kinases and transcription factors, among others, that control cell proliferation, differentiation, survival, and motility (Cargnello and Roux 2011). In physiologic conditions, these phenotypes are under tight molecular control by MAPKs, but in pathologic conditions such as cancer, cardiovascular disease, and neurodegeneration, MAPK signaling is often aberrant.

Although MAPKs encompass a large number of kinases, the best known MAPKs are ERKs (1 and 2), p38s (α, β, γ, and δ), and JNKs (1, 2, and 3). MAPKs control cell response programs by phosphorylating and thereby regulating the activity of many proteins implicated in senescence. In particular, ERK1/2 regulates senescence-associated proteins including RSKs, Sprouty, and MYC (Campaner et al. 2010; Macia et al. 2014; Munoz-Espin and Serrano 2014; Sun et al. 2018) and p38 regulates ATF6, ZNHIT1, HBP1, p53, MK2, and MK5 (Zhang et al. 2006; Sun et al. 2007; Debacq-Chainiaux et al. 2010; Druelle et al. 2016; Macedo et al. 2018). Besides regulating transcription of senescence-associated genes, MAPKs and their effectors (e.g., MNK1, MK2, RSKs) can also control gene expression programs post-transcriptionally by phosphorylating and thereby modulating the activity of RNA-binding proteins (RBPs) implicated in senescence, such as HuR, AUF1, PTBP1, TTP, GRSF1, and hnRNPA1 (Wang et al. 2005; Wang et al. 2016; Ziaei et al. 2012; Alspach et al. 2014; Wiley and Campisi 2016; Georgilis et al. 2018; Noh et al. 2018; Noh et al. 2019). Phosphorylation by MAPKs often alters the ability of RBPs to bind target mRNAs, as shown for HuR, TTP, and AUF1, and modulates the fate of these mRNAs (Grammatikakis et al. 2017; Soni et al. 2019). In addition, the MAPK substrates MK2, MNK1, and RSK appear to be essential for the translation of SASP factors (Herranz et al. 2015; Roux and Topisirovic 2018; Sun et al. 2018) and link MAPKs with the mTOR pathway, which is activated in senescent cells (Tomimatsu and Narita 2015). In short, MAPK signaling governs transcriptional and translational programs in senescent cells.

MAPKs as sensing elements in senescence

As a cellular response to sublethal damaging or oncogenic stressors, senescence relies on MAPKs to implement robust and specific molecular programs. Traditionally, activation of ERK1/2 has been associated with mitogenic signaling, while p38 and JNK are generally implicated in stress signaling (Dhillon et al. 2007; Pimienta and Pascual 2007). In senescence, the ERK1/2 pathway is activated by aberrant RAS/RAF signaling, DNA damage, and oxidative stress (Pearson et al. 2001; Kim et al. 2003; Khalil et al. 2011). RAS activates the RAF-MEK axis to induce ERK1/2 phosphorylation (Mebratu and Tesfaigzi 2009), while sublethal genotoxic damage signal through ATM-AKT to activate ERK1/2 and implement an anti-apoptotic, pro-senescence program (Viniegra et al. 2005; Bozulic et al. 2008; Khalil et al. 2011). Of note, senescence-associated oxidative damage has been shown to activate ERK1/2 through the inhibition of MAPK phosphatases (Kim et al. 2003) as well as by the induction of signaling of receptor tyrosine kinases (RTKs) by reactive oxygen species (Lei and Kazlauskas 2009).

Oxidative stress is traditionally linked to p38 activation (Xu et al. 2014). In this regard, ERK1/2 signaling triggered by oncogenes (e.g., RAF(V600E) or HRAS(V12G)) induces senescence at least partly through a rise in oxidative stress leading to the activation of p38 (Wang et al. 2002). However, only a few mechanisms by which oxidative stress directly activates p38 to trigger cell senescence have been proposed (Passos et al. 2010; Kodama et al. 2013), and other pathways, such as DNA damage-induced senescence through increased GADD45A expression by p53 have implicated p38 as a contributor to the response (Moskalev et al. 2012).

Senescence cell cycle arrest by MAPKs

The indefinite growth arrest of senescent cells is established and maintained primarily by the cyclin-dependent kinase (CDK) inhibitors p21 (CDKN1A), p16 (CDKN2A), and p15 (CDKN2B) (Hernandez-Segura et al. 2018; Casella et al. 2019). The abundance of these CDK inhibitors is under the direct control of MAPKs, particularly ERK1/2 and p38. Paradoxically, while ERK1/2 signaling promotes cell proliferation in dividing cells, it promotes growth arrest in senescent cells; as reported, altered translocation to the nucleus as well as magnitude and duration of activation redirected ERK1/2 function towards cell cycle arrest instead of proliferation (Ebisuya et al. 2005; Zou et al. 2019).

In senescence, p21 is induced transcriptionally by p53 following DNA damage (d’Adda di Fagagna 2008). The transcriptional status of p53 is regulated post-translationally in several ways, including phosphorylation, altered protein levels, and subcellular localization (Baar et al. 2017; Hafner et al. 2019). p38 directly influences p53 activity in senescent cells by phosphorylating p53 at Ser15 and thus stabilizing it (Xu et al. 2014), while the p38-effector MK5 phosphorylates p53 and enhances transcription of p21 mRNA (Sun et al. 2007). ERK1/2 also modulates p53 post-translationally (Wu 2004), but these effects are not well studied in senescence. MAPKs enhance p21 mRNA transcription through transcription factors activated by stimuli other than DNA damage (Abbas and Dutta 2009); for instance, both p38 and ERK1/2 activate ELK1 and SP1, which drive transcription of p21 mRNA (Shin et al. 2011; Kim et al. 2014b). In addition, ERK1/2 activation promotes p21 transcription by SP1 and SMAD proteins (Pardali et al. 2000; Kim et al. 2006; Luo 2017). Thus, even in the absence of DNA damage, MAPKs strongly elevate p21 abundance. Accordingly, ERK1/2 activation contributes to developmental senescence, a senescence phenotype that relies mainly on DNA damage-independent induction of p21 (Munoz-Espin et al. 2013; Storer et al. 2013).

The senescence proteins p16 (CDKN2A) and p14 (ARF) are expressed from the CDKN2A locus (Munoz-Espin and Serrano 2014); p16 inhibits CDKs that phosphorylate RB, while p14 helps stabilize p53 (Kim and Sharpless 2006). Transcription of the CDKN2A locus is repressed epigenetically through Polycomb group (PcG) proteins (Bracken et al. 2007; Ito et al. 2018). In this paradigm, the MAPK effector MK3 phosphorylates and reduces the levels of PcG protein BMI1, thus promoting senescence (Voncken et al. 2005; Lee et al. 2016). Additionally, transcription from the CDKN2A locus is controlled by SWI/SNF protein complexes (Kia et al. 2008), which evict PcG proteins and enhance p16 transcription. In this context, MAPK p38 positively regulates the function of the SWI/SNF protein BAF60 (Simone et al. 2004). Furthermore, p38 facilitates the transcription of p16 mRNA by activating the histone acetyltransferase P300 (Li et al. 2010; Wang et al. 2012). Finally, transcription of p16 mRNA is further promoted by MAPKs that activate ETS, SP1, and MSK1 (Ohtani et al. 2001; Wu et al. 2007; Shin et al. 2011; Culerrier et al. 2016).

MAPKs also modulate the activity of RBPs that control the stability and/or translation of mRNAs encoding senescence-associated CDK inhibitors. In this context, MNK1 phosphorylates hnRNPA1 and dissociates it from p16 and p14 mRNAs, rendering them more stable and enabling increases in p16 and p14 protein levels (Zhu et al. 2002; Ziaei et al. 2012). In another example, phosphorylation of HuR by p38 increases HuR binding to p21 mRNA, increasing p21 mRNA stability and elevating p21 levels (Wang et al. 2000; Lafarga et al. 2009), even though HuR levels decline overall in senescent cells (Wang et al. 2001; Lee et al. 2018). TTP phosphorylation by the MAPK effector MK2 leads to dissociation of TTP from p21 mRNA and increases p21 mRNA stability and p21 production (Al-Haj et al. 2012). Finally, degradation of the RBP AUF1 by the proteasome in an MK2-regulated manner might contribute to the stabilization of target p21 and p16 mRNAs and the reduction in telomerase transcription seen in senescent cells (Wang et al. 2005; Chang et al. 2010; Pont et al. 2012; Li et al. 2013).

Regulation of SASP by MAPKs

The SASP is a complex trait believed to be responsible for many of the pathophysiologic effects of senescent cells (Gorgoulis et al. 2019). SASP factors include many proinflammatory cytokines, growth factors, angiogenic factors, and matrix metalloproteinases.

MAPKs are upstream regulators of NF-κB, a major transcriptional coordinator of the SASP. Upon senescence-inducing stimuli, p38 enhances the DNA damage-driven NF-κB transcriptional activity, which in turn promotes the transcription of SASP genes including IL6, IL8, and GM-CSF (Rodier et al. 2009; Freund et al. 2011; Alimbetov et al. 2016). Although not assessed in senescent cells, MSK1, an effector of p38 and ERK1/2, enhances NF-κB function and increases the transcription of SASP factors IL6 and CXCL8 (Vermeulen et al. 2003; Reber et al. 2009). In senescence induced by oncogenic RAS, elevated ERK1/2 signaling promoted NF-κB-mediated SASP protein production (Catanzaro et al. 2014). Activation of the MAPK substrate RSK1, an enhancer of protein synthesis, elevated IL8 production (Sun et al. 2018), while the MAPK substrate MNK1 phosphorylated eIF4E and thereby enhanced the translation of proteins including SASP factors and MK2 (Wendel et al. 2007; Wu et al. 2013; Herranz et al. 2015). Activated MK2, in turn, phosphorylated ZFP36L1 and thereby suppressed its ability to degrade target mRNAs encoding SASP components (Herranz et al. 2015). Finally, a recent report shows that JNK activation in senescent cells promotes cGas-STING signaling and enhances the SASP (Vizioli et al. 2020).

Among the many SASP factors regulated independently of NF-κB (Davalos et al. 2010), TGFβ, PDGFA, and CTGF were induced by NOTCH signaling in senescent IMR-90 fibroblasts, producing a distinct early wave of the SASP (Hoare et al. 2016). TGFβ promotes senescence by increasing the expression of p15 and p21 (Munoz-Espin et al. 2013), and activates p38 through TGFβ-activated kinase 1 (TAK1) (Yu et al. 2002; Passos et al. 2010). Conversely, p38 can induce TGFβ production by activating NOTCH signaling in RAS-overexpressing fibroblasts or by activating ATF2 (Kim et al. 1992; Weijzen et al. 2002). mTOR is required for the induction of TGFβ in senescent cells (Aarts et al. 2017), but whether MAPKs modulate the early SASP is still unknown.

MAPK-regulated RBPs also play important roles in SASP regulation. Beyond influencing growth arrest (see above), RBPs such as AUF1, HuR, hnRNPA1, GRSF1, and TTP are also linked to the expression of SASP factors (Ross et al. 2012; Hashimoto et al. 2014; Alspach and Stewart 2015; Wang et al. 2016; Noh et al. 2019). For example, MNK1 phosphorylated hnRNPA1 (Ziaei et al. 2012) and enabled NF-κB transcription of proinflammatory SASP mRNAs (Wang et al. 2016), while phosphorylation of ZFP36L1 by MK2 caused dissociation and stabilization of IL8 or IL1B mRNAs, encoding major SASP factors (Herranz et al. 2015). Activated MK2 might also phosphorylate AUF1 to induce its dissociation from target IL6 and IL8 mRNAs in senescent cells (Alspach and Stewart 2015). The reduction of HuR levels in senescence increased the levels of NF-κB-regulated SASP factors (Hashimoto et al. 2014), while other RBPs, such as PTBP1, induced the SASP trait globally (Georgilis et al. 2018). MAPK-regulated RBPs that specifically promote SASP factor production could be promising therapeutic targets.

MAPKs in senescent cell survival

Senescent cells are normally resistant to apoptotic cell death, even though they show activation of apoptotic pathways. To persist within tissues, senescent cells rely on pro-survival nodes, including proteins in the anti-apoptotic programs driven by BCL2, PI3K-AKT, and p53 (Childs et al. 2014; Baar et al. 2017; Kirkland and Tchkonia 2017; Yosef et al. 2017). In light of evidence that clearance of senescent cells from tissues improves age-related pathologies (including fibrotic diseases, sarcopenia, cardiovascular disease, cachexia, diabetes, and Alzheimer’s disease (van Deursen 2014; McHugh and Gil 2018)), there is intense interest in exploiting senolysis in aging, age-related diseases, and situations in which senescent cells accumulate (e.g., cancer therapy).

While elevated levels of cell damage promote apoptosis, sublethal doses lead cells to a senescent state (Childs et al. 2014). In this respect, MAPK signaling together with other signaling pathways such as the DNA damage-p53 axis contribute to implementing an appropriate cellular outcome (Gong et al. 2010). For example, low doses of DNA damage cause ERK1/2 activation, ensuring cell survival, but higher doses fail to activate ERK1/2 and instead result in cell death by apoptosis (Dai et al. 2008; Khalil et al. 2011), and in some cases, ERK1/2 activation may itself promote cell death (Tang et al. 2002). Activation of p38 generally increases with damage severity (Gong et al. 2010; Lumley et al. 2017), but p38 may also be pro-apoptotic or pro-survival depending on the trigger and cell type (Igea and Nebreda 2015); in some cases, the pro-survival effect of p38 in senescent cells may be linked to the induction of autophagy (Slobodnyuk et al. 2019). Additionally, p21 induction opposes apoptosis in senescent cells, and its expression is enhanced by both ERK1/2 and p38 MAPKs (see above) (Yosef et al. 2017). A deeper understanding of how MAPK signaling in senescent cells influences apoptosis is needed.

One of the most frequent approaches to eliminate senescent cells from aged tissues and organs is by treating with a combination of Dasatinib plus Quercetin (D + Q). The senolytic effect of this combination of drugs was first discovered in a screen of proliferative and senescent cells (Zhu et al. 2015). Quercetin, a flavonoid, is a pleiotropic inhibitor of many kinases (Russo et al. 2012; Russo et al. 2017; Bruning 2013). It can promote apoptosis in response to different stresses that activate ERK1/2 or p38 and induce senescence (Kim et al. 2014a; Gong et al. 2018). Dasatinib inhibits many tyrosine kinases (Montero et al. 2011) including SRC, BCR-ABL, C-KIT, PDGFR, and Ephrin receptors (Daulhac et al. 1999; Matsumoto et al. 1999; Kim et al. 2008; Galan-Moya et al. 2008; Abbaspour Babaei et al. 2016; Kania and Klein 2016), and thereby blocks signaling through MAPKs. Interestingly, some ligands linked to the function of these tyrosine kinases are SASP factors, and thus Dasatinib might be suppressing the anti-apoptotic protection engendered by SASP factors. Moreover, the receptor-associated protein Caveolin-1 is highly expressed on the senescent cell plasma membrane (Volonte and Galbiati 2009; Kim et al. 2017) and can help activate receptors by secreted factors such as TGFβ, Ephrins, EGF, and FGFs (Pol et al. 2000; Razani et al. 2001; Vihanto et al. 2006; Katz et al. 2007; Shao et al. 2008; Feng et al. 2012).

Maintenance of senescence by MAPKs

Identifying the mechanisms by which the senescent phenotype is maintained long-term is a major question in aging and cancer. MAPKs are proposed to contribute to this persistent phenotype, as activation of both p38 and ERK1/2 increases over time in senescence (Kim et al. 2003; Freund et al. 2011). Indeed, oxidative stress caused by the persistent activation of ERK1/2 can preclude the function of phosphatases, thus creating a positive feedback loop (Kim et al. 2003; Colavitti and Finkel 2005). Additionally, in senescent cells ERK1/2 and p38 promote the expression of Caveolin-1, which interacts with and inactivates phosphatases PP2A or PP2C, thus retaining active MAPKs and ATM and reinforcing the constitutive signaling through MAPKs and p53 (Meskiene et al. 2003; Dasari et al. 2006; Volonte and Galbiati 2009). The coordinated actions of p53 and p38 may contribute to the enduring growth arrest of senescent cells, since activation of p53 in response to MDM2 antagonists causes irreversible growth inhibition only under atmospheric oxygen (21% O₂), which can occur in certain instances of senescence such as during wound repair and in lung disease (Parrinello et al. 2003; Wiley et al. 2018). Given the role of p38 in sensing oxidant stress, persistent p38 signaling could promote a low but prolonged induction of p53 levels in senescence (Sun et al. 2007; Purvis et al. 2012).

MAPKs may also contribute to the maintenance of senescence by alternative ways. For example, senescent cells implement global chromatin rearrangements called senescence-associated heterochromatic foci (SAHF) (Narita et al. 2003). These foci form in cells that are engaged in growth arrest by pRB (Zhang et al. 2005) and require the p38 effector protein HBP1 (Zhang et al. 2013). The p38 MAPK pathway also fuels a DNA damage-dependent activation of NF-κB-STAT3, triggering a positive feedback loop in senescent cells (Kuilman et al. 2010; Freund et al. 2011. Finally, p38 and ERK1/2 may reinforce mTOR signaling for a complete, long-term implementation of the senescence phenotype (Leontieva and Blagosklonny 2010; Gutierrez-Uzquiza et al. 2012; Laberge et al. 2015).

Together, these findings suggest that MAPKs contribute to the maintenance of senescence and the long-term growth arrest phenotype of senescent cells.

MAPKs and senescence in age-associated brain diseases

Recent studies have revealed that cellular senescence plays a key role in many age-associated neurodegenerative pathologies, such as Alzheimer’s disease (AD) (Bussian et al. 2018; Zhang et al. 2019) and Parkinson’s disease (PD) (Chinta et al. 2018). Although MAPKs have not been tightly linked to senescence in neurodegeneration, evidence for their implications is beginning to emerge.

Phosphorylation of p38 is increased in brains from aged mice (Li et al. 2011), and p38 activity is increased in AD when compared with age-matched brains (Hensley et al. 1999). The fact that p38 phosphorylation increases in regions where neurofibrillary tangles are found (Zhu et al. 2000) helps to support a role for p38 in Tau phosphorylation (Hanger et al. 2009; Maphis et al. 2016). Interestingly, senescent astrocytes, which are central players in AD pathogenesis (Salminen et al. 2011; Bhat et al. 2012), are highly dependent on p38 signaling for their phenotype (Mombach et al. 2015). ERK1/2 activity has also been implicated in Tau phosphorylation and in enhancing neurodegeneration in AD (Ferrer et al. 2001; Kirouac et al. 2017; Sun and Nan 2017). Conversely, it was recently found that amyloid oligomers induce ERK1/2 pathway in the hippocampus in the early stages of AD (Faucher et al. 2015).

Cellular senescence has also been found to be detrimental in the pathogenesis of PD, as removal of senescent cells was protective in mice that developed PD after paraquat treatment (Chinta et al. 2018). PD pathogenesis could be exacerbated by SASP factors secreted by senescent astrocytes and microglia in PD brain (Calabrese et al. 2018). Although the role of MAPKs in PD-associated senescence has not been studied directly, MAPKs are central mediators of the cellular response to stress, inflammation, and survival signals, all of which occur in the PD environment. Moreover, the activation of p38 by stress factors in microglia leads to the proinflammatory environment that exacerbates neurodegeneration in PD (He et al. 2018). ERK1/2 activation is required for harmful astrocytic inflammation within the nervous system (Zhuang et al. 2005) and suppression of ERK1/2 improves some side-effects of PD treatment, such as L-DOPA-induced dyskinesia (Santini et al. 2007). In sum, more comprehensive knowledge of MAPKs implicated in neurodegeneration-associated senescence is needed.

Concluding remarks and perspectives

MAPK pathways are integral to senescent traits. In response to stimuli capable of triggering senescence, MAPKs function as sensors to identify the type and extent of damage, and help to decide whether the ensuing response ought to be cell proliferation, differentiation, apoptosis, senescence, or other. If cells adopt a senescent response, MAPKs participate directly in the various traits of senescence. First, MAPKs contribute to implementing the gene expression programs that enable indefinite growth arrest, including increasing production of p21 and p16. Second, as integral mediators of the SASP trait, MAPKs control the production and secretion of SASP factors transcriptionally via NF-κB-dependent and -independent pathways, as well as post-transcriptionally by controlling the activity of RBPs that govern the production of SASP and other senescence-associated factors. Third, MAPKs are central to the anti-apoptotic phenotype that ensures the long-term survival of senescent cells. Therefore, understanding the role of MAPKs in balancing senescence-associated proliferation, gene expression, and survival is essential for the design of effective senotherapies.

As MAPKs control a large number of downstream effectors, in-depth analysis is also needed for identifying the specific contribution of MAPKs to traits like growth arrest, SASP, and resistance to apoptosis. Along with these needs, there are still many aspects of senescence that remain to be fully understood. For example, superior senescence markers must be identified; at present, classical markers such as senescence-associated β-Gal activity, p21 or p16 abundance, and SASP factor levels are not universal markers of senescence, and they lack sufficient specificity and sensitivity. Given that senescence can be detrimental or beneficial depending on the specific senescence paradigm, identifying the contribution of MAPKs to each senescence phenotype could uncover new tools (e.g., MAPK inhibitors) for therapeutic benefit. Improved animal models of senescence must also be developed and studied, and full catalogs of proteins and RNAs driving human senescence programs in cultured cells and in tissues/organs in vivo must be identified systematically. Finally, robust panels of senolytic and senostatic interventions to eliminate or reprogram senescent cells, respectively, must be uncovered. As we gain deeper knowledge of the pleiotropic signaling pathways and gene expression programs driving senescence, including those orchestrated by MAPKs, we will be able to design more rational approaches to modify the senescent cell compartment for therapeutic benefit.

Funding information

This work was supported in its entirety by the NIA IRP, NIH.