Renal fibrosis is the common end point for all progressive kidney diseases, which leads to kidney failure by an excessive accumulation of extracellular matrix (Boor et al., 2010). Accumulating evidence suggests that senescence of kidney tubule cells influences kidney fibrosis (Schafer et al., 2018). Senescence is a special form of permanent cell cycle arrest, which limits cellular proliferative life span. Some senescent cells can be cleared by immune cells, termed acute (short-term) senescent cells, while chronic (long-term) senescent cells keep accumulating and creating early senescence events, finally aggravating the pathology (Munoz-Espin and Serrano, 2014; Kobbe, 2019). Different hallmarks of senescence have been recognized which are involved in diverse signaling pathways, including apoptosis markers, senescence-associated secretory phenotype (SASP) factors, and cyclin-dependent kinase inhibitors (Hernandez-Segura et al., 2018).

Senescent cells show resistance to apoptosis (Childs et al., 2014) and accumulate dysfunctional mitochondria (Korolchuk et al., 2017). Mitochondrial outer membrane permeabilization (MOMP) is responsible for apoptosis in numerous cell death pathways (Chipuk et al., 2006). In the intrinsic apoptosis pathway, Bcl-2 and the caspase family proteins play important roles (Van Opdenbosch and Lamkanfi, 2019; Ngoi et al., 2020). The Bcl-2 family is divided into three main groups: anti-apoptotic (Bcl-2, Bcl-xl, and Mcl-1), pro-apoptotic (Bax and Bak), and pro-apoptotic BH3-only (Bim, Bid, Bad, and Puma) (Anantram and Degani, 2019). Senescent cells are known to be in a primed apoptotic state, triggered by the abnormal upregulation of anti-apoptotic and pro-apoptotic proteins (Fan et al., 2020). Caspase family proteins are downstream players of MOMP in the intrinsic apoptosis pathway (Shalini et al., 2015), and after the activation of the Bax-Bak-dependent MOMP, cytochrome C is released from the mitochondria stimulating caspase-9 activation and its downstream executioners, caspases-3 and -7, to initiate apoptosis (Van Opdenbosch and Lamkanfi, 2019).

Cell cycle arrest is another typical characteristic of senescent cells, which is largely mediated through activation of either one or both p53/p21^CIP1/WAF1 (p21) and p16^Ink4a (p16)/pRb pathways (Kumari and Jat, 2021). p53/p21 is activated during DNA damage response, resulting in a p21-dependent G0/G1 cell-cycle arrest (Ceccaldi et al., 2012; Ou and Schumacher, 2018). On the other hand, p16 inactivates Retinoblastoma 1 (pRb) thereby inhibiting the action of the cyclin dependent kinases, leading to G1 cell cycle arrest (Rayess et al., 2012). Both p53/p21 and p16/pRb pathways are independent in senescence induction. Acute DNA damage causes a cell cycle arrest via the p53/p21 pathway, while chronic DNA damage followed by the induction of the p16/pRB pathway maintains cell cycle arrest and senescence (Sperka et al., 2012). As a key mediator of cell-cycle arrest, p21 also shows a p53-independent upregulation according to some research (Zhang et al., 2011; Ruan et al., 2020). Furthermore, p53 is also involved in the apoptosis process, as described previously (Rufini et al., 2013; Ou and Schumacher, 2018).

SASP factors are related to a DNA damage response and are generally proinflammatory and/or profibrotic compounds including numerous cytokines (e.g., IL-6 and IL-8), growth factors (e.g., TGF-β and CTGF), chemokines (e.g., CCL2), and matrix-metalloproteinases (e.g., MMP-1 and MMP-3) (Hernandez-Segura et al., 2018; Birch and Gil, 2020). Those proteins induce or maintain senescence through different pathways, contributing to kidney fibrosis (Docherty et al., 2019). Current efforts are focused on clearing senescent cells as a treatment option for prevention of kidney fibrosis development and progression. Senolytics represent a good option as they can selectively eliminate senescent cells participating in senescence associated pathways by interfering with anti- and pro-survival signaling (Zhu et al., 2015). However, suitable cell models are required to evaluate senescence development in kidney tubule epithelial cells and their response to senolytics.

We previously developed a conditionally immortalized proximal tubule epithelial cell line overexpressing the organic anion transporter 1 (ciPTEC-OAT1) and applied it successfully for pharmacological and toxicological investigations, including drug disposition and interaction studies (Wilmer et al., 2010; Nieskens et al., 2016). OAT1 is a first step in the elimination of organic anions in humans and is responsible for the uptake of many anionic (waste) products in kidney proximal tubules (Pou Casellas et al., 2021). Since the expression of OAT1 is rapidly lost when culturing (primary) PTEC in vitro, OAT1 was stably expressed in ciPTEC by lentiviral transduction. This cell line now allows prediction of drug-induced nephrotoxicity and drug-drug interactions of organic anions in vitro (Nieskens et al., 2016). OAT1 is also involved in the uptake of uremic toxins, known to participate in the uremic syndrome typical of chronic kidney disease (Nigam and Bush, 2019). Since senescence is a key factor contributing to chronic kidney disease, uremic toxins might play a role in this process as well.

CiPTEC was created by means of a temperature sensitive mutant U19tsA58 of SV40 large T antigen (SV40T) and the essential catalytic subunit of human telomerase (hTERT), to keep the characteristics of primary cells (Wilmer et al., 2010). Temperature-sensitive SV40T allows cells to proliferate at the permissive low temperature of 33°C but induces a proliferation block that resembles senescence at a non-permissive temperature of 37°C (Larsson et al., 2004; Wilmer et al., 2010). The hTERT maintains telomere length, preventing replicative senescence induced by telomere shortening (Bodnar et al., 1998). Some studies already showed a relation between senescence and SV40T conditional models, because both pRb and p53 are activated by SV40T at the non-permissive temperature leading to a senescence-like arrest in the cells (Larsson et al., 2004; Brookes et al., 2015). Therefore, we hypothesized that ciPTEC-OAT1 exhibits a senescence phenotype when cultured at non-permissive temperatures that can be used to study senescence development in kidney tubule epithelial cells and their response to senolytics. In the present study, we evaluated apoptosis markers and other common senescence markers in ciPTEC-OAT1 cultured at permissive and non-permissive temperatures at different time points, to investigate whether these cells can be implemented as a valid in vitro model to study kidney senescence. Day 0 in both groups is considered as the non-senescence group (control). Finally, the senolytics response was detected by means of cell viability assessment and senescence-associated β-galactosidase (SA-β-gal) activity.

Cellular senescence is an irreversible condition with cell cycle arrest, SASP, and apoptosis resistance, which contribute to chronic kidney disease, leading to fibrosis (Stenvinkel and Larsson, 2013; Hernandez-Segura et al., 2018). We previously developed ciPTEC-OAT1 to be used in drug screening and nephrotoxicity studies (Nieskens et al., 2016). In the present study, we demonstrate that the cell model is also suitable for studying tubular senescence in the kidney. We detected differential expression of apoptosis-associated markers, common senescence markers, and some typical SASP factors suggesting that ciPTEC-OAT1 obtains a senescence-like phenotype when cultured at a non-permissive temperature for 9 d. Furthermore, senescent cells appeared sensitive to senolytic drugs.

In the present study, we demonstrated that the ciPTEC-OAT1 cultured at a non-permissive temperature expressed common senescence markers. In particular, the decrease in Lamin B1 (Shimi et al., 2011) and upregulation in p21 (Calcinotto et al., 2019) have been described as characteristic features that are involved in maintaining senescence phenotype by regulating JNK and caspase signaling (Yosef et al., 2017). Further, increased SASP factors have been reported as proinflammatory and matrix-degrading molecules (Childs et al., 2015), including PAI-1 (Sun et al., 2019), IL-1β (Shi et al., 2019), CTGF (Jun and Lau, 2017), and IL-6 (Mosteiro et al., 2018). Another feature entails apoptosis, responsible for cell turnover and maintaining extracellular environment. For instance, the Bcl-2 family members modulate the delicate balance between pro- and anti-apoptosis (Ngoi et al., 2020). The upregulations of Bcl-2 and Bcl-xl suggest that ciPTEC-OAT1 became anti-apoptotic upon maturation in both conditions, whereas culturing at 37°C speeds up the process and induces a more prominent senescence phenotype. Senescent cells show downregulation of Mcl-1 on protein levels (Lee et al., 2018), which is consistent with our findings. Furthermore, the effectors Bax and Bak shuttle between cytosol and mitochondrial outer membrane with different rates (Pena-Blanco and Garcia-Saez, 2018), which might explain why we observed a differential expression in the proteins. BH3-only proteins bind to the BH3 domain of the anti-apoptotic Bcl-2 proteins via hydrophobic interactions, thereby promoting cellular apoptosis (Anantram and Degani, 2019). Of BH3-only proteins, Bid showed higher protein levels at different time points in the maturation group. The increasing trend in all other proteins points toward a priming of cells to undergo apoptosis, but the execution of the death program is restrained. These findings are in line with previous reports showing that following senescence induction by ionizing radiation, senescent cells upregulate pro-apoptosis markers (Chang et al., 2016; Baar et al., 2017). Therefore, cellular senescence in our model with an upregulation of pro-apoptotic markers despite having an anti-apoptotic phenotype argues for cells searching for a new balance to maintain homeostasis.

Caspases are another group of proteins involved in cell death mediated by apoptosis and important senescent markers (Shalini et al., 2015). After MOMP, caspase activation takes place often within minutes, leading to cell death (Pena-Blanco and Garcia-Saez, 2018). Inhibition of caspases therefore blocks apoptosis. Here we detected that the activator (procaspase-9) and executioner (procaspase-3 and procaspase-7) were upregulated in a permissive temperature group but show a differential pattern when cells are cultured at the non-permissive temperature. When mitochondrial-mediated apoptosis is induced and caspase-9 and caspase-3 are activated, the expression of Bax and Bcl-2 has been reported to show different levels to maintain their balance (Zhang et al., 2021). Irradiation-induced senescence is accompanied by an upregulation of procaspase-3, -7, and -9 (Chang et al., 2016; Baar et al., 2017) but a downregulation of activated caspase-3 (Chang et al., 2016), in line with our results. The final downregulation of procaspases observed in our study might be explained by the cleavage of procaspase-9 finally to active caspase-3 and caspase-7. But because of senescence induction and an adapted balance in Bcl-2 family proteins, procaspase-9 is inhibited, finally leading to apoptosis-resistance (Figure 7).

The observed loss of LaminB1 and upregulation of β-gal, well-known indicators of senescence (Hernandez-Segura et al., 2018), in ciPTEC-OAT1 cultured at a non-permissive temperature further confirms the phenotypical changes. Silencing LaminB1 immediately leads to inhibition of proliferation and the induction of senescence (Shimi et al., 2011). Our previous research regarding cell cycle analysis of ciPTEC-OAT1 at permissive and non-permissive temperatures (Mihajlovic et al., 2019) has shown that ciPTEC-OAT1 when cultured at 37°C for 1 or 7 d exhibits significantly reduced proliferation (less cells in the S phase) and an increased number of cells in the G0/G1 phase of the cell cycle, indicating halted proliferation at a non-permissive temperature. This is in line with our current results. The activity of the lysosomal β-gal reflects increased metabolic activity and enhanced lysosomal content typical of senescent cells (Hernandez-Segura et al., 2018). Transfection with the temperature sensitive SV40T gene allows the cells to become conditionally immortalized (Wilmer et al., 2010). Although downregulation of SV40T at 37°C allows activation of p53 and pRb, factors involved in both p53/p21 and p16/pRb pathways, there was no significant difference in their mRNA expression (Supplementary Figure S1) and p16 protein expression appeared undetectable. This suggests that senescence of ciPTEC-OAT1 may not be induced by the p16/pRb pathway. On the other hand, p53 is pivotal in determining the fate of the cells, implying that the p53/p21 pathway is key in the initiation of senescence (Mijit et al., 2020). Previously published works described a decline of total-p53 levels in stress-induced senescence in mice (Feng et al., 2007) and p21’s role in maintaining senescence in mice (Yosef et al., 2017), which is also in accordance with our in vitro data showing a decline in p53 levels and an increase in p21 levels at 37°C. Therefore, our results suggest that ciPTEC-OAT1 cultured at 37°C promotes senescence through the p53/p21 pathway.

In addition to previously tested markers, SASP factors are also important players in senescence. IL-6 maintains senescence through the p53/p21 pathway (Effenberger et al., 2014; Li Y. et al., 2020), shared by IL-8, which is expressed as a function of IL-6 (Kuilman et al., 2008). TGF-β1 and CTGF are other SASP factors reported to mediate senescence (Jun and Lau, 2017; You et al., 2019). CTGF is a downstream mediator of TGF-β1 and is regulated by TGF-β1 (Ou et al., 2020). TGF-β1 induces kidney fibrosis by accumulation of extracellular matrix and CTGF expression by activation of Smad3 and p53 (Li X. et al., 2020). Meanwhile, both CTGF and TGF-β1 induce senescence and are accompanied with the upregulation of IL-6 and IL-8 (Jun and Lau, 2017; Fan et al., 2019). Although not significant, our results show an increasing trend of both TGF-β1 and CTGF in 37°C group compared to 33°C group. PAI-1 is a major TGF-β1/p53 target gene in kidney fibrosis and is known to be elevated in senescent cells, correlating with increased tissue TGF-β1 levels (Samarakoon et al., 2019; Rana et al., 2020). In our study, PAI-1 and CTGF increased remarkably on the mRNA level, suggesting cellular senescence. TNF-α is another SASP factor and inducer of senescence (Guo et al., 2019), but our results showed no important differences over time and between the culture conditions, indicating the senescence of ciPTEC-OAT1 is not induced or maintained by TNF-α. There are some discrepancies between mRNA and protein levels of the obtained results. Despite being difficult to explain, these discrepancies might be due to differences in the regulation of transcription and protein translation processes, as well as protein turnover rate. Taken together, the results of the mRNA and protein levels, we believe that the cells exhibit a senescence phenotype at a non-permissive temperature.

Finally, the senolytics navitoclax, dasatinib, and quercetin were evaluated. Our results suggest that ciPTEC-OAT1 cultured at a non-permissive temperature was sensitive to senolytics, and the Bcl-2 family inhibitor navitoclax (Tse et al., 2008; Chang et al., 2016), navitoclax, appeared most effective in selectively reducing viability of cells presenting senescent phenotype. A clinical trial of the dasatinib and quercetin cocktail demonstrated a decrease in p21 and p16 positive human adipose tissue cells and plasma SASP factors of diabetic kidney disease participants (Hickson et al., 2019). Senolytics treatment of ciPTEC-OAT1 led to a dose-dependent reduction of SA-β-gal positive cells, in line with previous results (Zhu et al., 2015). It has been suggested that SA-β-gal activity may be an outcome rather than a cause of senescence (Lee et al., 2006; Piechota et al., 2016), and our findings argue for a clearance of senescent cells leading to a reduction in total cell number. Our follow-up research will focus on investigating the underlying mechanisms of senolytics used in this study.

In conclusion, our results suggest ciPTEC-OAT1 can be used as a valid proximal tubule cell model both for mechanistic studies inherent to kidney senescence and fibrosis and for senolytic effects of newly developed drugs and their combinations.