Acute kidney injury (AKI) is a common and devastating pathologic condition in part due its higher incidence in the elderly and its association with an increased short- and long-term mortality (Levey and James, 2017; Hounkpatin et al., 2019; Martin-Cleary et al., 2019; Logan et al., 2020). Moreover, AKI is closely related to chronic kidney disease (CKD) as AKI may accelerate CKD progression to end-stage renal disease (ESRD) and CKD predisposes to AKI (Venkatachalam et al., 2015; Siew et al., 2016; Ruiz-Ortega et al., 2020). All these facts underscore the importance of the research in this area. Furthermore, the cellular and molecular mechanisms of the increased sensitivity to AKI in elderly patients are incompletely understood (Mehran et al., 2019; Infante et al., 2020; Aleckovic-Halilovic et al., 2021), hampering the design of any preventive or therapeutic approaches.

Kidney tubular cells comprise the bulk of the kidney cell mass and may be injured by hypoxia, toxic compounds, metabolic disorders and proteinuria, among other factors. In response to an insult, tubular epithelial cells undergo phenotype changes associated with tubular function impairment and activation of inflammatory, fibrotic and cell death pathways, which may reflect a state of cellular senescence (Linkermann et al., 2014; Ruiz-Ortega et al., 2020). The initial phase of AKI is followed by a recovery phase characterized by activation of protective and regenerative mechanisms that restore epithelial properties and functions in surviving cells (Yang et al., 2010). Tubular cell death in AKI can involve several cell death pathways, such as apoptosis and regulated necrosis (Linkermann and Green, 2014). Cells dying by regulated necrosis release intracellular molecules, called damage-associated molecular patterns (DAMPs), which amplify the inflammatory response by the activation of neutrophils and other immune cells in a process termed necroinflammation. There are several forms of regulated necrosis, including necroptosis, ferroptosis, and pyroptosis (Newton and Manning, 2016). Necroptosis, the best-characterized form of regulated apoptosis, is elicited by the binding of the receptor-interacting protein 1 (RIPK1) to RIPK3, leading to its oligomerization and autophosphorylation. Then, the active RIPK1-RIPK3 complex (also called necrosome) activates the pseudokinase mixed lineage kinase domain-like protein (MLKL), which translocates to the cellular membrane, causing cell membrane permeabilization, rupture, and subsequent cell death (Newton and Manning, 2016). Necroptosis plays an important role in experimental AKI, as described in renal ischemia/reperfusion injury (IRI), folic acid (FA)-AKI and cisplatin nephropathy (Linkermann et al., 2013; Linkermann and Green, 2014; Xu et al., 2015; Martin-Sanchez et al., 2017; Martin-Sanchez et al.,2018a). Ferroptosis, a caspase-independent cell death pathway, is characterized by reduced glutathione activity or content, reduced glutathione peroxidase 4 (GPX4) protein levels, massive lipid peroxidation and cell loss (Martin-Sanchez et al., 2020). Targeting ferroptosis by chemical inhibition or gene expression modulation reduced tubular injury and improved renal function in different experimental models, including IRI and FA-AKI (Martin-Sanchez et al., 2017; Martin-Sanchez et al., 2020).

Cellular senescence represents a maladaptive response to AKI, characterized by prolonged cell-cycle arrest (Melk et al., 2004; Bonventre, 2014; Gorgoulis et al., 2019). Following an initial insult, DNA damage activates a protective mechanism consisting in the arrest of the cell cycle and the activation of the DNA damage response (DDR) to facilitate DNA repair. After successful DNA repair, cells re-enter the cell cycle (Branzei and Foiani, 2008). Nevertheless, persistent activation of this protective mechanism can contribute to damage, as observed in disease conditions associated with cellular senescence (Gire and Dulic, 2015). Regarding the kidney, accumulation of senescent (in particular tubular epithelial) cells has been implicated in regeneration failure and AKI-to-CKD transition (Schmitt and Cantley, 2008; O’Sullivan et al., 2017; Kim et al., 2019). In this sense, prolonged tubular epithelial cell-cycle arrest, sustained inflammation, and impaired mitochondrial function can contribute to CKD progression (Levey and James, 2017; Andrade et al., 2018; Sato and Yanagita, 2018; Jiang et al., 2020). Cellular senescence or premature aging in the kidney is characterized by increased expression of some cell-cycle-related molecules such as the cyclin kinase inhibitors p16ink4a, p21cip, and p53 (Melk et al., 2004; Andrade et al., 2018; Knoppert et al., 2019; Koyano et al., 2019). Senescent cells are also characterized by a detrimental secretome known as senescence-associated secretory phenotype (SASP) (Melk et al., 2004; Bonventre, 2014). This secretome is enriched with pro-inflammatory cytokines, growth factors and profibrotic proteins such as IL-6, TGF-β, CTGF/CCN2 and PAI-1 (Acosta et al., 2013; Zhou et al., 2020) and is able to spread the senescence phenotype to neighboring cells (paracrine senescence) (Acosta et al., 2013), and to promote kidney fibrosis (Ferlicot et al., 2003; Melk et al., 2004; McGlynn et al., 2009; Yang et al., 2010; Günther et al., 2017; Valentijn et al., 2018). While inflammation is one of the first steps in tissue repair, persistent inflammation contributes to CKD progression (Cao et al., 2015; Rabb et al., 2016; Sato and Yanagita, 2018). Cytokines and interleukins within the SASP contribute to enduring inflammation and to further tubular cell injury and dysfunction (Kirkland and Tchkonia, 2017). Moreover, a low-grade inflammatory milieu is known to be present in the aged tissues, a condition named “inflammaging” (Greene and Loeser, 2015; Rea et al., 2018). “Immunosenescence” is a related concept, in which the dysfunctional immune response in the elderly presents characteristics related to cellular senescence and promotes inflammation, thus playing a crucial role in inflammaging (Sato and Yanagita, 2019; Schroth et al., 2020).

In the present study, we sought to elucidate the potential cellular and molecular mechanisms contributing to the increased severity of AKI in old age. To this aim, we have investigated whether aging-related processes, such as induction of cellular senescence, inflammaging and loss of renal protective factors, can modulate tubular damage, including cell death pathways activation and phenotype changes induced by AKI. Previous experimental studies have reported age-related exacerbation of renal injury in different AKI models. Now, we have investigated the FA-AKI model that presents a different mechanism of kidney injury (crystalluria with intratubular obstruction) than those in prior studies (cytokine storm) or on exogenous (cisplatin) or endogenous (heme) molecules that are directly toxic to tubular cells (Maddens et al., 2012; Nath et al., 2013; Wen et al., 2015). Aging is a process that has no fixed start date and does not occur suddenly. Rather, human glomerular filtration rate starts decreasing progressively from age 18–24 years (Wetzels et al., 2007). Most previous experimental studies on AKI and aging used mice from 15 to 18 months old (Maddens et al., 2012; Nath et al., 2013; Wen et al., 2015). However, in a sepsis AKI model an increase in mortality was already observed at 12 months (Maddens et al., 2012). Therefore, this time point was chosen to investigate early age-associated changes that could be responsible for increased AKI susceptibility.

The studies done in the murine model of AKI by exposure to the toxic compound FA revealed increased acute tubular damage in aging mice. Similarly, more severe drug-related AKI effects in elderly subjects have been described in humans (Metz-Kurschel et al., 1990; Khan et al., 2017), supporting the relevance of this experimental model to explore kidney disease (Miyauchi, 1991; Fan et al., 2017; Montgomery et al., 2017). The characterization of acute tubular damage at molecular level reveals an exacerbation of the tubular injury marker KIM-1 in old mice. The observations regarding the mechanisms triggered by AKI point to an age-related magnification of several proinflammatory-related processes, including gene overexpression of some proinflammatory factors (Lcn-2, Cxcl1, and Il-6), overactivation of inflammatory-cell-death pathways such as necroptosis, and amplification of cellular senescence including immunosenescence (Figure 10). Moreover, our experimental data, showing an exacerbation of renal damage in 12-months-old mice associated with the loss of renal protective factors, support the idea that age-associated susceptibility to AKI may start earlier than previously thought.

After an ischemic or nephrotoxic AKI insult, a wide range of pathophysiological events occur, including changes in tubular cell phenotypes, such as loss of physical cell–cell interactions and partial epithelial-to-mesenchymal transition (EMT) (Ruiz-Ortega et al., 2020), or even tubular cell death mediated by apoptosis and prominent programmed and unprogrammed necrosis (Martin-Sanchez et al., 2018b; Martin-Sanchez et al., 2018a). The proximal tubular cell is an important target of AKI (Chevalier, 2016), as we have confirmed in the FA model by de novo expression of the tubular damage biomarker KIM-1 in these cells. Some reports have found a lack of difference in initial severity of IRI, as described by no changes in tubular injury score, between aged and young mice after 1 day post IRI (Sato et al., 2016; Kim et al., 2019), whereas in our model of low-dose of FA-induced AKI, we found an exacerbated increase in KIM-1 positive tubular cells in old mice. Accordingly, in other murine AKI models induced by kidney exposure to toxic compounds, such as heme proteins (Nath et al., 2013), cisplatin (Wen et al., 2015) or bacteria inoculation (Maddens et al., 2012), a significant tubular damage in the acute phase was also described in old mice, showing age-related predisposition of tubular injury in response to toxic-induced damage.

Injured tubular cells are an important source of proinflammatory cytokines and chemokines, which contribute to the amplification of the inflammatory response (Liu et al., 2018). In this sense, FA-injected old murine kidneys presented a synergistic upregulation of proinflammatory genes, such as Lcn-2, Cxcl1, and Il-6, that codify cytokines and chemokines involved in the recruitment of infiltrating immune cells in the kidney. Previous experimental studies have also investigated the inflammatory response in the initial phase of AKI in old mice. In the IRI model, the number of macrophages significantly increased after 1 day in both young and aged mice (Kim et al., 2019), as found in other models of toxic exposure (Nath et al., 2013; Wen et al., 2015). These data clearly indicate that the combination of advanced age and exposure to toxics or ischemia induces an exacerbated innate inflammatory response in the injured kidney at this acute time point and suggest an increased susceptibility of the elderly to AKI. In addition, in the IRI model, exacerbation of immune response and changes in macrophage phenotypes is involved in the AKI-to-CKD transition (Kim et al., 2019).

Many evidences in humans indicate that the elderly exhibit low-grade systemic chronic inflammation even in healthy conditions (Goronzy and Weyand, 2013; Franceschi and Campisi, 2014). Moreover, aging-related dysregulation of several innate and acquired immune responses have been described (Goronzy and Weyand, 2013; Montecino-Rodriguez et al., 2013; Franceschi and Campisi, 2014), and in human kidney transplant patients, aged donor kidneys were observed to attract more infiltrating inflammatory cells than young ones (Øien et al., 2007). Inflammaging of the kidney has also been demonstrated by a microarray analysis of human samples (Rodwell et al., 2004). However, no significant renal changes were found at gene level for proinflammatory factors in healthy old (12 month) mice. Outstandingly, a magnification of the FA-AKI-induced pro-inflammatory response was observed in aging mice, which could be either a cause or a consequence of increased tubular damage. Among the proinflammatory mediators potentially involved in AKI exacerbation, IL-6 has special relevance. We have found that Il-6 gene expression was synergistically upregulated in FA-AKI in old mice, as previously described in a model of hemoglobin-induced AKI (Nath et al., 2013). Since IL-6 is a proinflammatory cytokine and a SASP component, targeting IL-6 or its downstream signaling could be an interesting therapeutic option in AKI in the elderly.

Tubular cell death is a feature of AKI and both apoptosis and regulated necrosis pathways are activated during FA-AKI (Sanz et al., 2008; Martin-Sanchez et al., 2017; Martin-Sanchez et al.,2018a). Caspase 3 activation is a central event in apoptosis (Justo et al., 2006; Linkermann et al., 2013; Martin-Sanchez et al., 2018a). In the present study, FA-AKI was associated with caspase 3 activation in young but not in old mice. In cultured tubular epithelial cells, inhibition of caspases is known to switch the mode of cell death induced by inflammatory cytokines from apoptosis to necrosis pathways (Justo et al., 2006; Martin-Sanchez et al., 2018a). In accordance with the latter, lack of caspase 3 activation in old FA-AKI mice was associated with evidence of involvement of the necroptosis pathway, i.e., RIPK3 and MLKL upregulation. This is a key difference to point out between young and old FA-AKI mice in our study, since apoptosis is a non-inflammatory form of cell death while necroptosis promotes inflammation (Martin-Sanchez et al., 2018b). In addition, ferroptosis is also overactivated in old FA-AKI, as shown by increased lipid peroxidation. Treatment with Ferrostatin-1, a ferroptosis inhibitor, prevented the inflammatory response and the expression of necroptotic proteins in FA-injected mice (Martin-Sanchez et al., 2017; Martin-Sanchez et al., 2018a), showing that this form of cell death is also related to inflammation. The observed downmodulation of apoptosis in old FA mice is also in line with the induction of a senescent phenotype of tubular cells in injured kidneys, since senescent cells are characteristically protected from apoptosis (Knoppert et al., 2019). These results are supported by the increased baseline and post-FA-induced AKI expression of the antiapoptotic protein BCL-xL showed in old, but not in young mice. In summary, our findings indicate an aging-related change in cell death mechanisms linked to increased tubular injury, characterized by an activation of proinflammatory cell death pathways (necroptosis and ferroptosis) and suppression of non-inflammatory cell death pathways (apoptosis) (Figure 10).

As mentioned above, regulated necrosis can also contribute to age-related amplification of AKI-induced renal inflammatory response. Thereby, DAMPs released by necrotic cells can produce innate immunity cell-derived cytokines by the activation of identical pattern recognition receptors, such as Toll-like receptors expressed on tissue-resident or infiltrating immune cells (Ujiie, 1989; Wu et al., 2007; Kurts et al., 2013). DAMP-associated inflammation is one of the earliest processes following AKI and contributes to an amplification of the loop of cell death/inflammation (Mulay et al., 2016b). Among the immune cells, some studies have demonstrated that macrophages actively participate in necroptosis (Linkermann et al., 2014; Mulay et al., 2016a). In this sense, in the model of IRI, the gene deletion of RIPK3 or MLKL reduced macrophage infiltration and NLRP3 inflammasome activation (Chen et al., 2018). Our results demonstrated that the number of infiltrating macrophages was significantly higher in old FA-injected kidneys associated with an overexpression of the necroptosis components RIPK3 or MLKL, supporting an exacerbation of necroptosis-macrophage inflammatory pathway in aging AKI mice.

Apart from kidney pathologies, necroptosis-mediated inflammation plays an important role in a variety of age-related diseases such as Alzheimer’s disease, Parkinson’s disease, and atherosclerosis (Royce et al., 2019). Some studies have found an association of age-related increase in DAMPs circulating levels, such as mitochondrial DNA or high mobility group protein B1 (Davalos et al., 2013; Pinti et al., 2014), with circulating proinflammatory cytokines (TNF-α, IL-6) in humans, suggesting that DAMPs might play a role in low-grade systemic chronic inflammation described in the elderly (Goronzy and Weyand, 2013; Franceschi and Campisi, 2014). In the same way, some experimental data support a relation between necroptosis and inflammaging. For example, accelerated aging Cu/Zn superoxide dismutase (Sod1) deficient mice that exhibit increased levels of circulating proinflammatory cytokines (Zhang et al., 2013; Deepa et al., 2019) had elevated MLKL protein and gene expression in adipose tissue at 9 months compared with age-matched wild type mice (Royce et al., 2019). Although we have found increased inflammatory cell infiltration in the old mice kidneys, the evaluation of key components of the necroptosis pathway, such as RIPK3 or MLKL, in 1-year old C57BL/6 mice showed no changes at gene and protein levels in healthy kidneys compared to young ones, suggesting that there is no age-related activation of necroptosis in our experimental conditions.

Cellular senescence may occur as a result of cell-cycle arrest due to increased expression of cyclin kinase inhibitors (Knoppert et al., 2019). Previous studies in different AKI models have described a rapid upregulation of p21cip1 expression in the early phase of AKI (Megyesi et al., 1998; Yu et al., 2005; Hodeify et al., 2011). Accordingly, we found increased expression of p21cip1 and p16ink4a in FA-AKI mice. Some studies have proposed that p21cip1 prevents DNA-damaged cells from entering the cell cycle by directly inhibiting CDK2 activity (Yu et al., 2005), thus avoiding cell death by necrosis or apoptosis (Megyesi et al., 1998). Indeed, p21cip1 knockout mice showed increased susceptibility to AKI mediated by ischemia or nephrotoxins (Megyesi et al., 1998; Megyesi et al., 2001; Nishioka et al., 2014). In contrast, the model of renal ablation in p21cip1 knockout mice presented diminished cell-cycle arrest, amelioration of renal dysfunction and lower interstitial fibrosis (Megyesi et al., 1999). On the other hand, renal p21cip1 is essential for the beneficial effects of renal ischemic preconditioning (Nishioka et al., 2014). Moreover, distinct types and severity of kidney injury can behave differently regarding cell-cycle arrest (Yang et al., 2010). Therefore, the functional consequences of p21cip1 expression are cell and disease context specific. In the present study, renal p21cip1 mRNA expression and tubular p21cip1 nuclear staining were significantly higher in old FA-induced AKI than in young mice. Furthermore, the DNA damage response marker γH2AX was also significantly activated in old AKI mice, showing mainly nuclear positive staining in tubular cells. Similarly, activation of prolonged cell-cycle arrest have also been reported in other experimental AKI models, but in this case, linked to fibrosis (Yang et al., 2010). In IRI-AKI mice, treatment with a p53 inhibitor has demonstrated the importance of G1 cell-cycle arrest in the progression of fibrosis (Lim et al., 2018). Another mechanism involved in senescence-mediated renal damage is related to the induction of SASP in injured tubular cells (Acosta et al., 2013). Here, we observed that in FA-induced AKI there was a significant increase in SASP gene expression (including Tgfβ1, Ctgf/Ccn2, Il6, and Serpine-1) in old mouse kidneys. Taken together, this data suggests that there is a magnification of the senescence phenotype in aged AKI mice (Figure 10). Interestingly, our results showed that in old murine injured kidneys, also some infiltrating immune cells were p21cip1 or γH2AX positive, suggesting molecular senescence in the immune cells in the aging kidney may be involved in the aggravated AKI response to FA in old mice (Figure 10). Although the exact cause of inflammaging is not known, cellular senescence (Campisi and D’Adda Di Fagagna, 2007) and immune senescence (Franceschi et al., 2000; McElhaney and Effros, 2009) have been proposed to play a key role in this process.

Finally, yet another remarkable finding was the reduced expression of nephroprotective factors Klotho and PGC-1α and their dramatic further downregulation induced by AKI in old mice. Klotho is normally expressed and secreted by tubular cells and has anti-aging, anti-inflammatory and anti-fibrotic properties (Kuro-o et al., 1997; Kurosu et al., 2005; Sanchez-Niño et al., 2013). Klotho downregulation can be both a consequence and driver of inflammaging in kidney disease (Moreno et al., 2011; Izquierdo et al., 2012; Fernandez-Fernandez et al., 2018; Sanchez-Niño et al., 2019; Fernández-Fernández et al., 2020). For example, Klotho protects endothelial cells from senescence (Carracedo et al., 2012). PGC-1α is the master regulator of mitochondrial biogenesis and PGC-1α deficiency is known to promote spontaneous kidney inflammation and to increase the severity of AKI (Fontecha-Barriuso et al., 2019; Fontecha-barriuso et al., 2020). Thus, the loss of the nephroprotective factors Klotho and PCG-1α due to ageing could contribute to an increased inflammatory and fibrotic response to FA-AKI.

In conclusion, our data indicate that aging kidneys lose local nephroprotective factors and reveal a switch to a proinflammatory cell death (necroptosis and ferroptosis) instead of apoptosis (Figure 10), associated to a synergistic upregulation of several proinflammatory (Lcn-2 and Cxcl1) and SASP mediators, such as IL-6. Moreover, these changes may partially underlie the age-related increased susceptibility to developing more severe AKI in response to toxic compounds, as clearly showed by a dramatic increase of KIM-1 expressing tubular cells (Figure 10). Another characteristic of severe AKI in aging kidneys includes the induction of cellular senescence in intrinsic renal cells and inflammatory cells. These features could interfere with the resolution of acute injury and favor the AKI-to-CKD transition. All these data point out the relevance of investigating the effects of senolytic drugs on cell-death pathways involved in AKI. Better understanding of inflammaging and immunosenescence could contribute to identifying prevention and/or intervention points to mitigate the structural and functional impairment of the kidneys in elderly people. Given the increasing frequency of AKI in the elderly, this information may help to come up with age-specific interventions to prevent or treat kidney injury in this age group.