Patients with chronic kidney disease (CKD) experience a progressive loss of kidney function over a period of months or years. Consequently, the diseased kidneys fail to metabolize and excrete toxic metabolites and organic waste solutes, which are normally removed by healthy kidneys. It is known that such retained toxic products, commonly referred to as “uremic toxins,” are transported by the blood to other organs, causing adverse effects throughout the body (D’Hooge et al., 2003; Meijers et al., 2010). In fact, toxic metabolites or organic solutes accumulated owing to CKD are known to be directly responsible for various complications, such as cardiovascular disease, anemia, pericarditis, renal osteodystrophy, and neurological disorders (Meijers et al., 2010; Gabriele et al., 2016). To date, more than 100 different types of uremic toxins have been found. These uremic toxins can be divided into two major categories: 1) free, water-soluble substances that reversibly bind serum proteins and can be removed by dialysis; and 2) protein-bound solutes and peptides that are difficult to remove by conventional dialysis, so they can stay in the body for a long period of time (Neirynck et al., 2013). p-Cresol (PC) belongs to the second type of uremic toxins, as it binds to proteins. PC is frequently detected in the serum of CKD patients (Faure et al., 2006). PC induces cell senescence, apoptosis and immunodeficiencies in the cardiovascular system, brain, and kidneys of affected patients (Azevedo et al., 2016). However, effects of PC on adult stem cells in patients with CKD have not yet been elucidated.

When cells age or become damaged by external stimuli, they are replaced by new cells that differentiate from adult stem cells into corresponding tissue-specific cell types to maintain tissue stability (Krause et al., 2001). Such adult stem cells are divided into various types: hematopoietic stem cells (Till and Mc, 1961), mesenchymal stem cells (Phinney and Prockop, 2007), and neural stem cells (Altman and Das, 1965). Among them, mesenchymal stem cells (MSCs) have a special self-renewal capacity with the potential to differentiate into multiple cell types, including osteoblasts, chondrocytes, and adipocytes, in multiple organs (Yun et al., 2009). MSCs are distributed all over the body, including bone marrow, cord blood, and adipose tissue; they are also present in large numbers in the bloodstream (Qiao et al., 2008). Owing to their ubiquitous nature, MSCs are easily exposed to uremic toxins that circulate in the blood of CKD patients. As a result, they may be vulnerable to toxin effects that can potentially complicate tissue regeneration process. Thus, it is important to evaluate possible harmful effects of uremic toxins, such as PC, on MSCs and find a way of controlling uremic toxin-induced MSC senescence.

Individuals with CKD experience disturbances of sleep, which may be linked to impaired endogenous melatonin secretion during nighttime (Maung et al., 2016). In light of known tissue regeneration problems in CKD patients, it is of interest to note that high blood melatonin levels may alleviate tissue regeneration disorders and other complications caused by CKD (Maung et al., 2016). Melatonin (N-acetyl-5-methoxy tryptamine), secreted by the pineal gland, is a hormone that regulates sleep and wakefulness (Hardeland et al., 2006). Recently, melatonin has been used clinically, and it is available off the counter as a sleeping aid that has a strong action and few side effects. The vast potential of melatonin as a therapeutic agent is explained by its ubiquitous nature. Melatonin acts not only on the brain, but also on various locations in the whole body, such as pineal gland, reproductive organs, skin, lymphocytes, and gastrointestinal tract (Acuna-Castroviejo et al., 2014). Specifically, melatonin is known to have many beneficial effects on the body, as it improves motility, induces proliferation of MSCs, and increases the therapeutic effect of MSC transplantation (Lee et al., 2014). Thus, in our study, we sought to investigate possible effects of PC on MSCs, and consequently evaluate the potential of melatonin as a therapeutic agent with protective properties against PC-induced stem cell senescence.

Cell culture

Human adipose tissue-derived MSCs were obtained from the American Type Culture Collection (Manassas, VA, USA). MSCs were confirmed to be pathogen- and mycoplasma-free. MSCs expressed specific cell surface markers [cluster of differentiation (CD) 73 and CD105, but not CD31]. MSCs were cultured in α-minimum essential medium (Gibco, Gaithersburg, MD, USA) supplemented with 10% (v/v) fetal bovine serum (Gibco), 100 U/mL penicillin, and 100 μg/mL streptomycin. MSC cultures were grown in a humidified atmosphere of 95% air and 5% CO₂ at 37°C.

Cell proliferation assay

MSCs were exposed to PC (0, 50, 100, 500 μM; Sigma-Aldrich, St. Louis, MO, USA) for 72 h or PC (500 μM; for 72 h) with preincubation melatonin (0, 1, 10, 100 μM; for 30 min; Sigma-Aldrich) or phosphate-buffered saline (PBS; vehicle). Cell proliferation parameters were assessed using a cell proliferation 5-bromo-2′-deoxyuridine (BrdU) ELISA colorimetric kit (Roche, Mannheim, Germany). To perform ELISA, 100 μg/mL BrdU was added to cultured MSCs and incubated at 37°C for 6 h. Then, 1 M H₂SO₄ was added to stop the reaction. Light absorbance of the samples was measured by an ELISA reader (BMG labtech, Ortenberg, Germany) at 450 nm.

Beta-galactosidase staining assay

Senescence was assessed by measuring the percentage of cultured cells that stained positively for senescence-associated beta-galactosidase (SA-β-gal) activity as described previously (Cong et al., 2006). Cells were washed with phosphate-buffered saline (PBS) twice and fixed with 2% formaldehyde/0.2% glutaraldehyde (Sigma-Aldrich). Fixed cells were incubated at 37°C (without CO₂) for 12 h with a β-gal staining solution (1 mg/mL X-Gal, 40 mM citric acid-sodium phosphate buffer, 150 mM NaCl, 2 mM MgCl₂, 5 mM potassium ferrocyanide, and 5 mM potassium ferricyanide; pH 6.0; Sigma-Aldrich). Stained (blue, positive) and non-stained (negative) cells were counted by using phase contrast microscopy (Nikon, Tokyo, Japan) in five independent cultures.

Catalase activity assay

MSCs were harvested from the culture dish by scraping with a rubber policeman on ice. To measure catalase activity, cell lysate protein extracts (40 μg) were incubated with 20 mM H₂O₂ (Sigma-Aldrich) in 0.1 M Tris-HCl (Sigma-Aldrich) for 30 min. Then, 50 mM Amplex Red reagent (Thermo Scientific, Waltham, MA, USA) and 0.2 U/mL of horseradish peroxidase (Sigma-Aldrich) was added and incub ated for 30 min at 37°C. Changes in absorbance values associated with H₂O₂ degradation were measured by an ELISA reader (BMG labtech) at 563 nm.

Western blot assay

MSCs homogenates were extracted by using RIPA lysis buffer (Thermo Scientific). Cell lysates (20 μg of total protein) were separated by using 6–15% sodium dodecyl sulfate polyacrylamide gel electrophoresis, and the proteins were transferred to nitrocellulose membrane (Thermo Scientific). After the blots had been washed with a solution comprising 10 mM Tris-HCl (pH 7.6; Sigma-Aldrich), 150 mM NaCl (Sigma-Aldrich), and 0.05% Tween-20 (Sigma-Aldrich), the membranes were incubated with 5% bovine serum albumin for 1 h at room temperature and then, incubated with appropriate primary antibodies against phosphoinositide 3-kinase (PI3K), phosphorylated PI3K (p-PI3K), Akt, phosphorylated Akt (p-Akt), phosphorylated 5′-AMP-activated protein kinase (AMPK), mechanistic target of rapamycin (mTOR), phosphorylated mTOR (p-mTOR), LC3, beclin 1, ATG7, p62/Sequestosome 1 (p62/SQSTM1), senescence marker protein 30 (SMP30), and β-actin (all from Santa Cruz Biotechnology, Dallas, TX, USA). The membranes were then washed more than twice, and the primary antibodies were detected using a goat anti-rabbit IgG or a goat anti-mouse IgG conjugated to horseradish peroxidase (Santa Cruz Biotechnology). The bands were visualized by enhanced chemiluminescence (Amersham Pharmacia Biotech, Little Chalfont, UK).

Dihydroethidium staining

To measure superoxide anion levels in cultured MSCs, the cells were incubated with 10 μM dihydroethidium (DHE; Sigma-Aldrich) for 30 min at 37°C. After washing with PBS three times, samples were visualized by fluorescence microscopy (Zeiss, Oberkochen, Germany) or analyzed by flow cytometry (Cyflow Cube 8 FACS; Sysmex Partec, Görlitz, Germany). Data were analyzed using standard FSC Express software (De Novo Software, Los Angeles, USA).

Statistical analysis

All data are expressed as the mean ± standard error of the mean (SEM). Results of all experiments were analyzed by one-way analysis of variance (ANOVA). Comparisons of more than three groups were made by using the Bonferroni-Dunn test. Differences were considered to be statistically significant if p<0.05.

Melatonin prevents the reduction in MSC proliferation rate and MSC senescence induced by PC

Exposure of MSCs to a range of PC concentrations (0–500 μM) for 72 h induced cell senescence as was indicated by decreased BrdU incorporation levels (Fig. 1A). At 500 μM PC, the lowest BrdU incorporation in MSCs was observed (Fig. 1A). It has been shown previously that treatment with 500 μM PC increased reactive oxygen species (ROS) production and inhibited cell viability, proliferation, and migration (Chang et al., 2014). To investigate the protective effect of melatonin, MSCs were pretreated with 0–100 μM melatonin before the exposure to PC for 72 h. We found that pretreatment with 100 μM melatonin reversed PC-induced decrease in BrdU incorporation (Fig. 1B), whereas incubation with lower concentrations of melatonin (1 and 10 μM) was without effect. Thus, 500 μM PC and 100 μM melatonin were used in the rest of the experiments to maximize the characteristic effects of these substances. To determine the effect of melatonin on cell senescence conferred by PC, cytochemical detection of SA-β-gal activity was performed. Melatonin was found to decrease PC-induced SA β-gal activity (Fig. 1C, 1D). These results support the notion that melatonin can counteract the reduction of MSC proliferation and induction of cell senescence caused by PC.

Melatonin inhibits PC-induced reactive oxygen species generation via PI3K/Akt-dependent catalase activation

The cytoprotective effect of melatonin is thought to be mediated via its binding to melatonin receptors and subsequent activation of various signaling pathways that promote cell survival (Lee et al., 2014; Yu et al., 2017). Thus, we sought to specifically examine whether melatonin affected PI3K/Akt-dependent catalase activity and inhibited ROS generation. As shown in Fig. 2A and 2B, melatonin time-dependently increased levels of phosphorylated PI3K and Akt. Based on these results, we hypothesized that activation of the Akt signal transduction pathway by melatonin could affect catalase activity and ROS generation. Indeed, melatonin increased catalase activity, and this effect was sensitive to an Akt inhibitor (Fig. 2C). Next, to establish whether melatonin could suppress ROS generation induced by PC, we incubated MSCs with DHE, a widely used small-molecule fluorescent probe specific for O₂⁻ (Zhao et al., 2003). The reaction between O₂⁻ and DHE generates 2-hydroxyethidium, a highly specific red fluorescent product, that shifts excitation and emission peaks from 350 and 400 to 518 and 605 nm, respectively (Zielonka et al., 2008). MSCs were pretreated with melatonin in the presence or absence of an Akt inhibitor for 2 h and then exposed to PC for 12 h. Changes in DHE fluorescence excitation and emission peak intensities were quantified by using confocal microscopy and flow cytometry. We found that incubation with PC increased DHE intensity and the number of DHE-positive cells. These changes were inhibited by melatonin (Fig. 2D–2F). These results suggest that melatonin reduced PC-induced ROS generation through increased catalase activity via Akt pathway activation.

Involvement of AMPK/mTOR pathway in the protective effect of melatonin

The increase in ROS is closely related to the activity of AMP-activated protein kinase (AMPK) (Li et al., 2015). AMPK regulates various cellular processes, such as cell growth, apoptosis, and autophagy by modulating mTOR activity (Li et al., 2015). To investigate the involvement of AMPK and mTOR in melatonin-induced cytoprotective effect, we determined changes in AMPK expression and phosphorylation of mTOR induced by the exposure to PC or combined incubation of PC with melatonin. We observed that 500 μM PC increased total AMPK expression (Fig. 3A, 3B), but decreased mTOR phosphorylation (Fig. 3A, 3B) in MSCs in a time-dependent manner. Melatonin completely reversed these effects of PC, whereas pretreatment with an Akt inhibitor, in turn, counteracted the changes caused by melatonin (Fig. 3C, 3D). These results suggest that the protective effect of melatonin on PC-induced MSC senescence is mediated by AMPK and mTOR signaling pathways.

Melatonin inhibits PC-induced autophagy via Akt signaling

Excessive autophagy is associated with the inhibition of cell proliferation and increase in the number of senescent cells (Yang et al., 2011). To examine the effects of melatonin on PC-induced autophagy, we first determined how PC modulated the expression of autophagy-related proteins, such as LC3, beclin 1, ATG7, and p62/SQSTM1, which are key molecules in autophagosome formation. PC time-dependently increased the expression levels of LC3, beclin 1, and ATG7, but decreased the expression of p62/SQSTM1 (Fig. 4A, 4B). Correspondingly, pretreatment with melatonin counteracted PC-induced changes in LC3, beclin 1, and ATG7 expression levels, and restored p62/SQSTM1 expression reduced by PC (Fig. 4C, 4D). Finally, as in the experiments on the modulation of total AMPK and p-mTOR levels, co-application of an Akt inhibitor neutralized the effects of melatonin (Fig. 4C, 4D). Taken together, these results suggest that melatonin potentially affects expression levels of autophagy-related proteins via activation of the Akt pathway, ultimately preventing the induction of autophagy by PC.

Protective effect of melatonin on PC-induced MSC senescence

Expression of senescence marker protein 30 (SMP30) decreases with age (Maruyama et al., 2010). In our experiments, PC decreased SMP30 expression in a time-dependent manner (Fig. 5A), which was in line with other senescence-related changes induced by PC. We found that pretreatment with melatonin rescued the reduction of SMP30 caused by PC, whereas protective effects of melatonin were, in turn, sensitive to the Akt inhibitor (Fig. 5B). As was shown above (Fig. 1C, 1D), melatonin prevented pro-senescence effects of PC. That effect of melatonin was also sensitive to the inhibition of the Akt pathway (Fig. 5C, 5D). These results suggest that melatonin diminished PC-induced reduction in the expression level of SMP30, which is involved in cell growth and anti-aging processes. This activity of melatonin was another manifestation of its potent restoring effects against PC-induced cell senescence.

In our study, we for the first time demonstrated that the uremic toxin PC, found in individuals with CKD, caused MSC senescence. Furthermore, we demonstrated that pretreatment with melatonin prevented PC-induced senescence of MSCs by activating the PI3K and Akt pathways and by inhibiting ROS-dependent, AMPK pathway-mediated autophagy.

PC, a metabolite of environmental toxins, is a uremic toxic solute normally excreted by the kidneys. In conditions such as CKD, PC accumulates in the body (D’Hooge et al., 2003). In CKD, uremic toxins chronically stimulate renal endothelial cells. PC decreases the expression of endothelial adhesion molecules of endothelial cells and promotes monocyte adhesion to endothelial cells, thereby increasing endothelial monolayer permeability and inflammation (Dou et al., 2002; Cerini et al., 2004). Actual concentration of PC and the mechanism of its formation in several kidney diseases, including glomerulopathy, tubulointerstitial nephropathy, and acute kidney injury, still remain unclear. PC has been shown to activate caspase signaling and induce apoptosis and necrosis of renal epithelial tubular cells, producing a characteristic phenotype of CDK-related damage (Brocca et al., 2013). In addition, PC induces renal proximal tubular cell death by triggering autophagy and apoptosis (Lin et al., 2015). When not excreted properly, uremic toxic solutes adversely affect not only the kidneys, which are their primary target, but also other organs and systems (Meijers et al., 2010; Gabriele et al., 2016). Oxidative stress, DNA/mitochondrial damage, apoptosis, and the number of senescent cells were reported to increase in cardiomyocytes of mice with modeled cardiorenal syndrome in vivo (Chua et al., 2016). In addition, treatment of H9c2 cardiomyocyte cells with PC resulted in a decrease in mitochondrial ATP levels and enhanced senescence (Chua et al., 2016). With regard to physiological functions of stem cells, it has been shown that insulin promoted the differentiation of bone-marrow derived mouse MSCs into target cells in vivo by upregulating expression levels of HIF-1α, VEGF, and stromal cell-derived factor 1α via the PI3K/Akt-dependent pathway (Noh et al., 2014). PC has been previously shown to inhibit this signaling pathway and suppress mouse MSC differentiation (Noh et al., 2014). Thus, we initiated our study assuming that PC effect on mouse cells could be translated to a human model, so that normal physiological functions, such as proliferation and senescence, would be similarly affected in human MSCs. Indeed, our results showed that incubation with PC inhibited BrdU incorporation and increased the number of β-galactosidase-positive cells (Fig. 1), indicating inhibition of proliferation and induction of senescence. In an attempt to prevent MSC senescence induced by PC, we chose to pretreat MSCs with melatonin, which is an easily applied, inexpensively synthesized, natural protector with few side effects. Melatonin is an endogenous hormone secreted by the pineal gland (Hardeland et al., 2006). Its physiological blood concentrations are maintained within pM or low nM range (Reiter and Tan, 2003). In some tissues, melatonin is present at considerably higher concentrations because of tissue specificity (Reiter and Tan, 2003), and it has a precise concentration-dependent effect on stem cells. Specifically, it has been reported that at low concentrations, proliferation, differentiation, and survival of stem cells was increased by melatonin (Fu et al., 2011), whereas at higher melatonin concentrations, stem cell pluripotency became notably lower (Zhang et al., 2010). Because of this variable concentration-dependent effect on target cells, it was extremely important to find a precise concentration of melatonin that could prevent PC-induced senescence. In our study, we revealed that 100 μM melatonin restored MSC proliferation reduced by PC and partially counteracted the increase in the number of β-galactosidase positive cells, whereas at concentrations of 1 μM and 10 μM, melatonin did not have any significant effect (Fig. 1). Thus, we concluded 100 μM was the optimal concentration of melatonin for our experiments that sought to address possible signaling mechanisms of the protection against PC-induced MSC senescence.

Previous studies suggested that PC induced oxidative stress in CKD patients by activating catalase and NADPH oxidase (Watanabe et al., 2013; Chang et al., 2014), leading to cell cycle arrest and cytotoxicity. Melatonin, in turn, has been ascribed anti-oxidant properties and demonstrated to alleviate cell damage (Yu et al., 2017) via activation of pro-survival and proliferation-related signaling, such the Akt pathway (Mehrzadi et al., 2016). Consistent with these findings, we found that catalase activity and ROS generation were significantly increased in MSCs exposed to PC and that pretreatment with melatonin prevented these changes. In addition, pretreatment with an Akt inhibitor blocked melatonin rescuing effect against PC action, suggesting that melatonin-induced PI3K/Akt pathway activity is essential in the prevention of PC-induced catalase activation, ROS generation, and, ultimately, MSC senescence. Nevertheless, low concentrations of ROS are known to be effective for cell protection and intracellular debris elimination (Diehn et al., 2009; Park et al., 2016a). In addition, we found that melatonin counteracted increased expression of total AMPK and reduced phosphorylation of mTOR induced by PC (Fig. 3). As in the experiments that examined catalase activity, inhibition of Akt neutralized melatonin inhibitory effects on PC-induced changes in the expression of AMPK and phosphorylated mTOR (Fig. 3C, 3D). Consistent with our results, ROS-mediated AMPK up-regulation has been shown to inhibit mTOR phosphorylation, and to suppress stem cell differentiation and self-renewal (Adams et al., 2016; Riva et al., 2016). In addition, evidence suggests that the AMPK/mTOR signaling axis is important for autophagy-mediated cell senescence and aging (Michalik and Jarzyna, 2016; Ganesan et al., 2017). Despite that knowledge, PC has not been previously shown to modulate AMPK expression through the generation of ROS, and the possibility that this signaling can regulate mTOR phosphorylation has not been elucidated. Therefore, our results provide a new perspective on the mechanism of cell senescence induction by PC, implicating AMPK/mTOR signaling in this effect. Moreover, we showed that effects of PC on MSCs could be ameliorated by melatonin in an Akt pathway-dependent manner.

Next, we attempted to uncover consequences of mTOR dysregulation for other cellular responses that are important in cell senescence. PC has been shown to induce autophagy in human renal proximal tubular cells (Lin et al., 2015), whereas mTOR signaling is known to regulate autophagy during stem cell senescence and disease (Menendez et al., 2011; Liang et al., 2016; Maiese, 2016). However, whether PC can induce autophagy in MSCs and whether melatonin can prevent that effect, has not been clearly elucidated. It has been shown that PC could participate in autophagosome formation and regulate expression of autophagy-related proteins, such as ATG4, beclin 1, and LC3B-II, ultimately increasing autophagy response (Lin et al., 2015). In contrast, melatonin directly inhibited mTOR-dependent autophagy and ER stress-dependent autophagy, preventing autophagic cell death and senescence (Kang et al., 2014; Fernandez et al., 2015). In the present study, we observed that PC increased the expression levels of the autophagy marker proteins LC3, beclin-1, and ATG7 in MSCs, but decreased the expression of p62/SQSTM1. In addition, we found that preincubation with melatonin partially or completely restored changes in the expression of autophagy-related proteins altered by PC. Thus, it is plausible that protective effects of melatonin are partially mediated via the suppression of PC-induced autophagy. Furthermore, we found that pretreatment with an Akt inhibitor prevented the effect of melatonin, which highlighted the critical nature of Akt activation in downstream melatonin signaling. Melatonin has been previously characterized as an autophagy inhibitor that downregulates expression levels of autophagic proteins in myoblast and glioma cells (Kim et al., 2011, 2012; Martin et al., 2014). Therefore, we suggest that melatonin reduces autophagy and prevents MSC senescence induced by PC via activation of the Akt pathway. In support of this notion, we demonstrated that melatonin affected the expression of SMP30 (Fig. 5). SMP30 expression levels decrease with age and therefore, it is a critical marker of biological maturation and premature aging caused by pathological processes (Abbas et al., 2017; Lesniewski et al., 2017). In our experiments, PC-treated MSCs showed increased numbers of SA-β-gal-positive cells and lower levels of SMP30 expression (Fig. 5). Pretreatment with melatonin suppressed these changes, by decreasing the number of SA-β-gal-positive cells and increasing SMP30 levels (Fig. 5B-5D). Although our results do not directly provide a link between autophagy and SMP30 levels, a previous report showed that SMP30 expression was reduced by autophagy (Park et al., 2016b). Thus, our results strongly support the notion that melatonin-dependent activation of the Akt pathway counteracts PC-induced ROS/AMPK pathway-mediated autophagy and reduction in SMP30 expression, ultimately ameliorating MSC senescence caused by PC.

Overall, our experimental results highlight a considerable therapeutic potential of melatonin anti-autophagic effects that reverse PC-induced MSC senescence. By showing that the protective action of melatonin against PC-induced MSC senescence is sensitive to the incubation with an Akt inhibitor, we demonstrated key significance of the Akt signaling pathway activation in melatonin effects. Therefore, our findings reveal a novel role of melatonin and Akt pathway in autophagy and in MSC senescence associated with CKD. Although our study was performed in vitro, the results obtained suggest a way to rescue developing MSC senescence in patients with CKD. Further studies are required to examine the therapeutic effects of melatonin in detail by using a mouse model of CKD.

The authors declare that they have no conflicts of interest.

This work was supported by National Research Foundation grant funded by the Korean government (NRF-2016R1D-1A3B01007727, NRF-2017M3A9B4032528). The funders had no role in the study design, data collection or analysis, the decision to publish, or preparation of the manuscript.