Skip to main content
NCBI home page
Redox Biology logoLink to Redox Biology
. 2023 May 20;64:102757. doi: 10.1016/j.redox.2023.102757

A superoxide-driven redox state promotes geroconversion and resistance to senolysis in replication-stress associated senescence

Le Luo a, Shazib Pervaiz b,c,d,e,f, Marie-Veronique Clement a,c,d,e,f,∗
PMCID: PMC10265521  PMID: 37285741

Abstract

Using S-phase synchronized RPE1-hTERT cells exposed to the DNA damaging agent, methyl methanesulfonate, we show the existence of a redox state associated with replication stress-induced senescence termed senescence-associated redox state (SA-redox state). SA-redox state is characterized by its reactivity with superoxide-sensing fluorescent probes such as dihydroethidine, lucigenin and mitosox and peroxynitrite or hydroxyl radical sensing probe hydroxyphenyl fluorescein (HPF) but not the hydrogen peroxide (H2O2) reactive fluorescent probe CM-H2DCFDA. Measurement of GSH and GSSH also reveals that SA-redox state mitigates the level of total GSH rather than oxidizes GSH to GSSG. Moreover, supporting the role of superoxide (O2.-) in the SA-redox state, we show that incubation of senescent RPE1-hTERT cells with the O2.- scavenger, Tiron, decreases the reactivity of SA-redox state with the oxidants’ reactive probes lucigenin and HPF while the H2O2 antioxidant N-acetyl cysteine has no effect. SA-redox state does not participate in the loss of proliferative capacity, G2/M cell cycle arrest or the increase in SA-β-Gal activity. However, SA-redox state is associated with the activation of NF-κB, dictates the profile of the Senescence Associated Secretory Phenotype, increases TFEB protein level, promotes geroconversion evidenced by increased phosphorylation of S6K and S6 proteins, and influences senescent cells response to senolysis. Furthermore, we provide evidence for crosstalk between SA redox state, p53 and p21. While p53 mitigates the establishment of SA-redox state, p21 is critical for the sustained reinforcement of the SA-redox state involved in geroconversion and resistance to senolysis.

Keywords: Superoxide, Peroxynitrite, Senescence, Geroconversion, Senolysis, Replication-stress

Graphical abstract

Image 1

Open in a new tab

Highlights

  • •

    Superoxide-driven redox state is associated with key features of replication-stress induced senescence.

  • •

    This redox sate has been termed senescence-associated redox state (SA-redox state).

  • •

    SA-redox state is characterized by reactivity with superoxide/peroxynitrite-sensing probes but not hydrogen peroxide reactive CM-H2DCFDA.

  • •

    SA-redox state is sensitive to the redox modulator Tiron but not N-acetyl cysteine.

  • •

    Interplay between SA-redox state, p53 and p21 promotes geroconversion and resistance to senolysis.

1. Introduction

Senescence is described as a state of irreversible cell cycle arrest in which the cells are metabolically active but unable to proliferate. Aside from the characteristic structural and biochemical features such as flat and enlarged morphology, an increase in lysosome numbers and senescence-associated β-Galactosidase (SA-β-Gal) activity, senescence is heralded by the acquisition of a secretory phenotype (Senescence Associated Secretory Phenotype; SASP) in which the cells secrete a slew of cytokines/chemokines and metalloproteases that alter tissue physiology via paracrine and/or autocrine signaling [1]. Initially regarded as an artefact induced by stress of cell culture, acquisition of the senescent phenotype is now well established as a genuine cell state and fate with dichotomous functional outcomes. On the one hand, senescence serves to limit tumorigenesis and facilitate wound healing, but on the other hand aberrant accumulation of senescent cells in a tissue is associated with various pathological states linked to organismal aging [2,3]. To that end, elimination of senescent cells has been shown to provide protection against age-associated pathologies and delays ageing [4,5].

Cellular senescence starts with growth arrest due to the activation of DNA Damage Response (DDR) kinase, ATR/ATM, leading to the stability of tumor suppressor p53 through its phosphorylation at Serine 15. Subsequent to growth arrest, senescent cells undergo extensive changes, which contribute to their progression into a deep senescent state through the activation of intracellular signaling networks that are proving to be more complex than originally envisaged [6]. In this regard, Dulic et al. demonstrated that, despite growth arrest being the critical event for the initiation of senescence, other aspects of the senescent phenotype appear to occur independently of cell growth arrest [7]. This notion of uncoupling growth arrest and the senescent phenotype was reinforced by the hypothesis that establishment of the senescent phenotype includes cell growth arrest and inappropriate activation of mTOR, indicated by sustained S6 phosphorylation, an event known as geroconversion [[8], [9], [10]].

Interestingly, while conventional wisdom has it that oxidative stress is a causal mechanism upstream of DNA damage induction, a dynamic feedback loop triggered by DDR was identified, which after a delay of several days locks the cell in a state of deep cellular senescence. An essential feature of this loop is mitochondrial dysfunction and production of oxidants in response to the sustained expression of cell cycle checkpoint protein p21. This loop was shown to be necessary and sufficient for the stability of growth arrest during the establishment of the senescent phenotype [11]. The proposed mechanism by which growth arrest is stabilized was that p21-dependent production of oxidants replenish short-lived DNA damage foci and maintain an ongoing DDR and/or accelerate telomere shortening through the production of single or double strand DNA breaks (DSB). Furthermore, similar to the activation of survival and growth signaling pathways in the absence of exogenous stimuli seen in senescent cells, our previous reports demonstrate that oxidants, in particular superoxide anion (O2.-), promote cell survival by activating growth factor-independent proliferation or via inhibiting apoptotic execution [[12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32]]. Therefore, considering the role of oxidants in maintaining the senescent phenotype and the activation of growth-factor signaling and inhibition of cell death by a O2.--driven redox state, here we questioned the existence of a senescence linked redox milieu and its association with key phenotypic features of cellular senescence outside of the oxidants’ ability to induce DNA damage. Using a model of replication stress-induced senescence in human retinal pigmental cells-1 (RPE1-hTERT), we present evidence to support the existence of a redox state associated with the senescent phenotype that we termed as senescence-associated redox state or SA-redox state. We demonstrate that SA-redox state is involved in the establishment of key phenotypic features of cellular senescence, including geroconversion and senescent cells resistance to senolysis. We also provide evidence for an intricate crosstalk between SA-redox state, p53 and p21. Our findings could have implications for the design and/or use of antioxidants that would not only prevent oxidation of macromolecules but also target signaling pathways activated upon the acquisition of deep cellular senescence associated with aging.

2. Results

2.1. Induction of cellular senescence in S phase synchronized RPE1-hTERT cells is associated with a distinct superoxide-driven intracellular redox state

RPE1-hTERT cells were synchronized with 4 mM thymidine for 24 h followed by 1 h incubation with either normal culture medium or medium containing the DNA damaging agent methyl methanesulfonate (MMS). After 1 h, MMS containing medium or control medium were replaced with fresh culture medium before various key features of the senescent phenotype were assessed. Cells exposed to MMS are referred to as SN cells and cells exposed to control medium as pro cells (Fig. 1A). While 24 h after the addition of fresh medium, SN cells are still arrested at the S-G2/M boundary, pro cells resume normal cell cycle progression (Fig. 1B). A significant decrease in the expression of the proliferation capacity-associated nuclear protein, Ki-67, is observed in SN cells (Fig. 1C), which is associated with enlarged and flattened morphology (Fig. 1D) and increased SA-β−Gal activity (Fig. 1E). Furthermore, a significant increase in IL-6 secretion (Fig. 1F), upregulation of the transcription factor TFEB (Fig. 1H) together with lysosomal proliferation evidenced by increased Lamp1 expression (Fig. 1G), decrease in the DNA damage repair protein Rad 51 and loss of Lamin B1 (Fig. 1K) are observed in SN cells compared to pro cells. SN cells also present with persistent DNA damage as shown by the detection of γH2AX and 53BP1 foci (Fig. 1I and J). Finally, the significant upregulation of p53 and p21, but not p16, supports the induction of p53-p21-dependent DDR pathway in SN cells (Fig. 1K). Together, these data establish that SN cells demonstrate key phenotypic features of cellular senescence.

Fig. 1.

Fig. 1

Fig. 1

Open in a new tab

DNA damaging agent MMS induces cellular senescence in S phase synchronized RPE-1-hTERT cells. A) Protocol used to generate senescent (SN) and control (pro) RPE-1-hTERT cells following thymidine synchronization. B) Cell cycle analysis of pro and SN cells 6hr, 24hr, and 48hr after the addition of fresh medium. C) Immunofluorescence staining of pro and SN cells nuclei with the proliferation maker Ki67, 5 days after the addition of fresh medium. Hoechst dye is used to identify nuceli. D) Morphology of SN and pro cells 5 days after the addition of fresh medium. Representative images of SN and pro cells are shown at 40X or 100X magnification. E) β-gal activity in SN and pro cells 5 days after the addition of fresh medium. F) Secretion of IL-6 by SN and pro cells at day 1, 2, 3 and 5 following the addition of fresh medium. Results are expressed as picogram/ml of IL-6 normalized to cell number. The data represent the mean ± SEM of at least three independent experiments. *p < 0.05 comparing SN to pro cells at each time point by two-tailed unpaired t-test. G) Immunofluorescence staining of pro and SN cells for Lamp1 5 days after the addition of fresh medium. Hoechst dye is used to identify nuclei. H) TFEB protein expression in cell lysates of pro and SN cells day 1, 2, 3 and 5 after the addition of fresh culture medium. I) γH2AX and J) 53BP1 immunofluorescence staining of SN and pro cells 5 days after the addition of fresh medium. Hoechst dye is used to identify nuceli. K) Rad51, Lamin B1, p53, p21, p16 proteins expression in cell lysates of pro and SN cells at day 1, 2, 3 and 5 after the addition of fresh medium.

Having established that SN cells are in a senescent state, we next set out to investigate the intracellar redox status of SN cells compared to pro cells. To do so, we made use of various oxidant-reactive fluorescent probes namely CM-H2DCFDA (DCF-DA) for the detection of non-specific oxidants including H2O2, DAF-FM diacetate (DAF) for the detection of nitric oxide (NO), HPF for the detection of hydroxyl radical (OH.) and peroxynitrite (ONOO−), dihydroethidium (DHE) to detect cytosolic O2.- and mitosox to detect mitochondria O2.-. SN cells and pro cells were incubated with each of the fluorescent probes and analyzed by flow cytometry 1 h, 5 h, 24 h, 48 h, 72 h and 120 h following the addition of fresh medium (Fig. 2A). While no significant difference in the fluorescence of the H2O2-reactive probe DCF-DA for up to 120 h (5 days) was detected when comparing SN and pro cells, a transient lower DAF fluorescence was observed in SN versus pro cells at 24 h and 48 h (Fig. 2A). In contrast, a significantly higher fluorescence intensity of HPF, DHE and mitosox in SN cells was first detected 72 h which continued to increase up to 120 h (5 days) (Fig. 2A). Taken together, these data support that SN cells have a distinct redox state compared to pro cells. Because this redox state is associated with a senescent phenotype we termed this redox state as senescence-associated redox state or SA-redox state.

Fig. 2.

Fig. 2

Open in a new tab

SN cells present with a distinct superoxide driven intracellular redox milieu compared to pro cells. A) Intracellular oxidants' reactivity assessed using the fluorescent probes DCF-DA, DAF, HPF, Mitosox and DHE in SN and pro cells 1 h to 120 h after the addition of fresh medium. Fluorescence level in SN cells is expressed as percentage of the fluorescence measured in pro cells. Data shown are the mean ± SEM of at least three independent experiments. *p < 0.05 comparing SN to pro cells at each time point by two-tailed unpaired t-test with Welch's correction. B) and C) SA-redox state reactivity to B) lucigenin and C) HPF in SN, SN cells incubated with Tiron (SN + Tiron) and SN cells incubated with NAC (SN + NAC). Bar charts represent percentage of the fluorescence intensity relative to the signal detected in SN cells. Data represent the mean ± SEM of at least three independent experiments. *p < 0.05 using two-tailed unpaired t-test with Welch's correction. D) GSH/GSSG ratio, E) GSH and F) GSSG in pro, SN, SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC). Bar charts represent fold increase relative to SN cells. The data are shown as mean ± SEM of at least three independent experiments. *p < 0.05 using two-tailed unpaired t-test with Welch's correction.

To gain further insight into the nature of the SA-redox state, we incubated SN cells with an established SA-redox state (Day 5 after the addition of fresh medium) in culture medium alone or culture medium containing the O2.- scavenger, Tiron, or the H2O2 scavenger NAC for another 3 days. SN cells incubated with Tiron will subsequently be referred to as SN + Tiron cells and SN cells incubated with NAC will be called SN + NAC cells. Lucigenin and HPF fluorescent probes were used to assess the effect of the two redox modulators on SA-redox state. Reactivity to lucigenin and HPF was significantly lower in SN + Tiron cells compared to SN cells, while a slight increase in lucigenin chemiluminescence and no difference in the reactivity with HPF was detected in SN + NAC cells (Fig. 2B and C). These data, together with the lack of reactivity with the H2O2 specific fluorescent probe DCF-DA as well as the absence of an effect of H2O2 specific anti-oxidant NAC on the SA-redox state argue in favor of intracellular O2.-, not H2O2, as the key oxidant associated with SA-redox state. To further support that H2O2 is not an oxidant associated with SA-redox state Fig. 2D shows the inability of NAC to influence GSH/GSSG ratio in SN cells (Fig. 2D). Moreover, a detailed analysis revealed that the GSH/GSSG ratio was significantly lower in SN cells compared to pro cells, however, the lower GSH/GSSG ratio was mostly due to the decrease in GSH rather than an increase in GSSG (Fig. 2E and F), as would be expected upon H2O2 driven redox stress. In addition, while in SN + Tiron cells a reduced levels of both, GSH and GSSG, are detected compared to SN cells, GSH is increased in SN + NAC cells without significantly affecting the level of GSSG (Fig. 2E and F). Together, the lower lucigenin and HPF reactivity as well as the decrease in total GSH in SN + Tiron cells compared to SN cells provides testimony that intracellular O2.- rather than H2O2 is mainly responsible for the generation of SA-redox state and that SA-redox state is involved in regulating total GSH rather than the oxidation of GSH to GSSG as seen by absence of the change in the ratio of oxidized to reduced glutathione (GSH:GSSG) when comparing SN and SN + Tiron cells.

2.2. SA-redox state is not linked to cell cycle arrest, loss of proliferative capacity or SA-β−Gal activity

Intrigued by the observation that SA-redox state is only detected 72 h after the addition of fresh medium in SN cells while SN cells are maintained in a S-G2/M arrested state (Fig. 2A), we questioned the role of SA-redox state in the maintenance of cell cycle arrest. Results show that incubation of SN cells with Tiron (SN + Tiron) or NAC (SN + NAC) neither escape S-G2/M cell cycle arrest nor overcome the decrease in Ki-67 detected in SN cells (Fig. 3A and B). Furthermore, the increase in SA-β-Gal activity, one of the commonly used markers to assess cellular senescence, is unaffected in SN cells (SN) incubated with Tiron (SN + Tiron) or NAC (SN + NAC) (Fig. 3C and D). On the other hand, pro cells resumed cell cycle, expressed Ki67 and did not express SA-β-Gal activity (Fig. 3A, B, 3C & 3D).

Fig. 3.

Fig. 3

Open in a new tab

SA-redox state is not linked to cell cycle arrest, loss of proliferation capacity or SA-β-Gal. A) Cell cycle profiles, B) expression of Ki67, C) SA-β gal activity and D) C12FDG fluorescence in pro and SN cells and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC). D) Percentage of C12FDG positive cells is shown relative to SN cells and as the mean ± SEM of at least three independent experiments. *p < 0.05 using two-tailed unpaired t-test.

2.3. SA-redox state regulates SASP profile in a NFκb-dependent manner

Another key phenotypic characteristic of senescent cells is the acquisition of SASP, which involves the secretion of growth factors, chemokines, matrix metalloproteinases and cytokines such as IL-6 and IL-8 [33,34]. In agreement with the senescent phenotype of SN cells, results show increased secretion of IL-6 and IL-8 by SN cells compared to pro cells (Fig. 4A and B). Notably, while the presence of O2.- scavenger Tiron (SN + Tiron) significantly reduced the amount of secreted IL-6 and IL-8 compared to SN cells, NAC treated cells exhibited significantly higher IL-6 and IL-8 secretion (Fig. 4A and B). Interestingly, while the higher level of IL-6 secretion is consistent with a higher level of IL-6 transcription and intracellular protein expression in SN cells compared to pro cells (Fig. 4A, C and 4E), no significant difference in IL-8 transcription and intracellular protein expression is seen between SN and pro cells (Fig. 4B, D and 4E). Moreover, while the lower level of IL-6 secretion upon scavenging O2.- (SN + Tiron) is consistent with a lower level of IL-6 transcription and intracellular protein expression compared to SN cells (Fig. 4C and E), no significant change in IL-8 transcription and intracellular protein expression is seen between SN and SN cells incubated with Tiron (SN + Tiron) (Fig. 4D and E). Similarly, the increase in IL-6 secretion detected in SN cells incubated with NAC (SN + NAC) compared to SN cells correlates with increased IL-6 mRNA and protein levels (Fig. 4A, C and 4E) while the increase in IL-8 secretion is not associated with an increase in IL-8 mRNA and protein expression (Fig. 4B, D and 4E). Together, these data highlight a role for SA-redox state in regulating the secretion of two important SASP factors such as IL-6 and IL-8, however, the SA-redox state dependent increase in secreted IL-8 appears to only be at the level of protein secretion.

Fig. 4.

Fig. 4

Open in a new tab

SA-redox state regulates SASP profile. A) IL-6 and B) IL-8 secretion levels in pro, SN and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC). Secretion level is expressed as fold relative to the secretion detected in pro cells after normalization to cell numbers. Data represent the mean ± SEM of at least three independent experiments. *p < 0.05 by two-tailed unpaired t-test. C) IL6 and D) IL-8 mRNA expression in SN cells, SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC) and pro cells using qPCR and expressed as fold relative to pro cells. The data represent the mean ± SEM of at least three independent experiments. *p < 0.05 using two-tailed unpaired t-test. E) Intracellular IL-6 and IL-8 protein levels in pro cells (pro), SN cells (SN) and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC). Band intensities of IL-6 and IL-8 were normalized to the loading control β-actin before being expressed as fold relative to the expression of the protein in SN cells. Values represent the mean ± SEM of a minimum of three independent Western blots.

Since NF-κB is the key transcription factor driving the transcription of genes that code for SASP factors in most models of cellular senescence [33,35], we next assessed the activation of NF-κB in pro and SN cells and evaluated the effect of Tiron or NAC. In unstimulated cells, NF-κB is sequestered in its inactive form in the cytoplasm by association with members of IκB inhibitor proteins, such as IκBα. Upon stimulation, IκBα is phosphorylated and subsequently degraded, resulting in the release and nuclear localization of NF-κB p65/p50 subunits to activate downstream transcriptional targets [36]. Our results confirm the activation of NF-κB during the establishment of cellular senescence as seen by a decrease in IκBα and nuclear translocation of p65 in SN cells compared to pro cells (Fig. 5A and B). However, the same decrease in IκBα protein and the translocation of p65 were not seen in the presence of Tiron (SN + Tiron; Fig. 5A and B). On the other hand, a further decrease in IκBα protein expression without a significant effect on the nuclear translocation of p65 was detected in the presence of NAC (SN + NAC) compared to SN cells (Fig. 5A and B). Importantly, the increased expression and secretion of IL-6 and IL-8 in SN cells were significantly blocked in SN cells incubated with the NF-κB inhibitor, JSH-23 [37] (SN + JSH) (Fig. 5C and D). It is worth noting that IL-6 and IL-8 secretion were not significantly different when comparing SN cells incubated with JSH (SN + JSH), Tiron (SN + Tiron) and JSH + Tiron (SN + Tiron + JSH) (Fig. 5C and D). The involvement of NF-κB transcriptional activation was further corroborated by the inhibitory effect of JSH-23 on Tiron-sensitive increase in mRNA levels of IL-6 in SN cells (Fig. 5E). In contrast, the inhibition of IL-8 secretion is not accompanied with a lower level of IL-8 mRNA (Fig. 5F), suggesting a transcription-independent effect of NF-κB in promoting the secretion of IL-8. It is worth pointing out that JSH-23 also prevented the increase in IL-6 and IL-8 secretion seen in SN cells incubated with NAC (SN + NAC + JSH), while inhibiting the increase in GSH using the inhibitor of the enzyme gamma-glutamyl-cysteine synthetase, BSO, had no effect on NAC-induced IL-6 and IL-8 secretion (Suppl. Fig. 1A, B, C). Together, these data support that SA-redox state contributes to the activation of NF-κB in SN cells, and while NF-κB activation regulates the transcription and secretion of IL-6, only the secretion of IL-8 is dependent on NF-κB. Moreover, our data also support that NAC activates NF-κB in GSH- and SA-redox state independent manner that is associated with the increase in IL-6 and IL-8 secretion.

Fig. 5.

Fig. 5

Open in a new tab

SA-redox state regulates SASP profile in NF-κB-dependent manner. A) IκBα protein expression in pro, SN and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC) cells. Band intensities of IκBα were normalized to the loading control β-actin before being expressed as fold relative to the expression of the protein in SN cells. Values represent the mean ± SEM of a minimum of three independent Western blots. B) nuclear-cytosolic expression of p65 protein in pro, SN, and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC) cells. Hsp90 and Lamin A/C are used as fractionation markers for cytosolic and nuclear fractions, respectively. Band intensities of p65 were normalized to the loading control β-actin. p65 protein levels in nuclear fraction are expressed as fold relative to the level of the protein detected in SN cells. Values represent the mean ± SEM of a minimum of three independent Western blots. C–F) SN cells and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC) were incubated with 25 μM of the NF-κB inhibitor JSH-23 (SN + JSH, SN + Tiron + JSH23 and SN + NAC + JSH) for an additional 3 days before C) secretion of IL-6, D) secretion of IL-8, E) IL-6 mRNA and F) IL-8 mRNA were assessed. Results are expressed as fold relative to control pro cells after normalization to cell number. The data represent the mean ± SEM of at least three independent experiments. *p < 0.05 comparing SN + Tiron or SN + JSH vs SN and #p < 0.05 using one-way ANOVA.

2.4. SA-redox state is involved in the upregulation of TFEB protein

It has been shown that senescent cells process their chromatin via an autophagy/lysosomal pathway that might contribute to the stability of the senescent phenotype [38]. The master transcription factor involved in lysosomes biogenesis is the Transcription Factor EB (TFEB) [39]. While a similar increased TFEB protein expression is seen in SN cells and SN cells incubated with NAC (SN + NAC) compared to pro cells, the level of TFEB protein in SN cells incubated with Tiron (SN + Tiron) is lower than the one seen in SN cells and close to the level seen in pro cells (Fig. 6A). Furthermore, highlighting a delink between TFEB protein expression and its level of transcription, no significant change in TFEB mRNA level was detected between pro, SN and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC) (Fig. 6B). Moreover, no significant change in the rate of TFEB protein degradation was detected in SN, SN + Tiron or pro cells upon performing cycloheximide chase (Fig. 6C and D). These data show that the increase in TFEB protein in SN cells is dependent on SA-redox state, however, the increase in TFEB protein is neither associated with increased TFEB transcription nor reduced protein degradation.

Fig. 6.

Fig. 6

Open in a new tab

SASP is involved in the increased expression of TFEB. A) TFEB protein level and B) TFEB mRNA were assessed in pro, SN and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC). mRNA levels are expressed as fold relative to pro cells. Data presented are the mean ± SEM of at least three independent experiments. C) Pro, SN and SN cells incubated with Tiron (SN + Tiron) were subjected to cycloheximide chase for 16, 24 and 48 h. D) Band intensities of TFEB protein expression were normalized to the loading control β-actin. Results are expressed as the mean of the fold relative to protein level at 0 h ± SEM of at least three independent experiments.

2.5. SA-redox state is associated with phosphorylation of p70S6K and S6

Another important hallmark of senescence is the activation of the growth kinase, mTOR, which is usually considered as a marker of geroconversion, the activity of which is assessed through the phosphorylation of the kinase p70S6K and the ribosomal S6 protein [8,40]. Our results show significantly reduced T389P of p70S6K as well as S235/236P of the S6 protein in SN cells incubated with Tiron (SN + Tiron) compared to SN cells (Fig. 7A and B), while no significant difference was observed upon incubation with NAC (SN + NAC) (Fig. 7A and B).

Fig. 7.

Fig. 7

Open in a new tab

SA-redox state is associated with p70S6K and S6 phosphorylation. A) p70S6K and p70S6k-pT389 and B) S6 and S6-pSer235/236 in pro, SN and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC). A) S6K and B) S6 protein phosphorylation levels are obtained relative to total protein level and calculated relative to the level detected in SN cells. The data represent the mean ± SEM of at least three independent experiments. Note that the β actin loading control in Fig. 7B is similar to Fig. 4E because the same cell samples and membranes were used.

2.6. Targeting SA-redox state enhances senescent cells sensitivity to senolysis

Besides the effect of SA-redox state on the phenotypical changes that characterize senescent cells, senescent cells have been shown to lose their viability when exposed to specific drugs, plant extracts, peptides or cell-clearing therapies known as senolytics [1,41]. Hence, we questioned the role of SA-redox state senescent cells' response to the senolytic agent, ABT-737 [41]. Incubation of SN cells with increasing concentration of ABT-737 leads to a significant decrease in cell viability and increased cleavage of the apoptotic marker caspase 3 while no effect is seen when pro cells are exposed to ABT-737 (Fig. 8A and B). Interestingly, ABT-737-dependent decrease in cell viability in SN cells (SN) is even greater when SN cells are incubated with Tiron (SN + Tiron) while no difference in cells’ sensitivity to ABT-737 is seen between SN cells incubated with NAC (SN + NAC) and SN cells (Fig. 8C). These data are corroborated by the detection of cleaved caspase 3 in SN + Tiron cells at lower concentration of ABT-737; 300 nM ABT-737 induced comparable levels of caspase 3 processing in SN + Tiron cells to that obtained with 900 nM ABT-737 in SN cells (Fig. 8E). Finally, in agreement with the senolytic-specific nature of ABT-737, exposure of pro cells, pro cells incubated with Tiron (pro + Tiron) or pro cells incubated with NAC (pro + NAC) to ABT-737 did not significantly affect cells viability (Fig. 8D). Together, these data confirm that ABT-737 specifically affects the viability of senescent cells and not proliferating cells, and highlight the involvement of SA-redox state in determining the sensitivity of senescent cells to senolysis.

Fig. 8.

Fig. 8

Open in a new tab

SA-redox state regulates senescent cells response to senolysis. A) pro and SN cells, C) SN, and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC) and D) Viability of pro cells (pro), pro cells incubated with Tiron (pro + Tiron) or NAC (pro + NAC) was assessed after 48 h exposure to ABT-737. Cells' viability is represented as percentage relative to cells grown in absence of ABT-737. Data shown are mean ± SEM of at least three independent experiments. #p < 0.05 using multiple t tests comparing SN to pro cells at the respective concentration of ABT-737; *p < 0.05 using multiple t tests comparing SN + Tiron as compared to SN cells at the respective concentration of ABT-737. B) pro and SN cells and E) SN, SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC) were exposed to ABT-737 for 48 h before levels of cleaved and full-length caspase 3 were assessed.

2.7. Crosstalk between SA-redox state, p53 and p21

Next, we investigated the relationship between SA-redox state and two key proteins involved in the establishment of the senescent phenotype, the tumor suppressor p53 and the cyclin-dependent kinase inhibitor p21. As expected, an increase in p53 and p21 is observed in SN cells compared to pro cells (Fig. 9A). Interestingly, upon incubation of SN (SN) cells with Tiron (SN + Tiron) the level of γH2AX and p53 are higher compared to SN cells. On the other hand, when SN cells are incubated with Tiron (SN + Tiron) the level of p21 is comparable to that seen in pro cells. Having observed that incubation of SN cells with Tiron decreases the level of p21 protein expression to the one seen in pro cells, we questioned if directly decreasing p21 protein expression using a specific p21 siRNA to directly decrease p21 protein expression would phenocopy the effect of Tiron in SN cells. Results show that, unlike the results obtained when SN cells were incubated with Tiron (SN + Tiron), the secretion of IL-6 and IL-8 (Fig. 9B) or the level of GSH, GSSG and GSH/GSSG ratio (Fig. 9C) were not affected upon gene knockdown of p21 (SN + sip21) compared to SN control cells transfected with a control siRNA (SN + siCo). On the other hand, similar to the results obtained in SN cells incubated with Tiron (SN + Tiron), knockdown of p21 (SN + sip21) exhibited increased sensitivity to ABT-737 compared to SN cells (Fig. 9D). Notably, knockdown of p21 in pro cells had no effect on the cell viability upon exposure to ABT737 (Fig. 9E). The increase in sensitivity to senolysis upon sip21 (SN + sip21) is corroborated by the ability of significantly lower concentrations of ABT737 in inducing caspase 3 activation compared to SN cells transfected with control siRNA (SN + siCo) (Fig. 9F). Furthermore, the increase in γH2AX and p53 and decrease in S6 phosphorylation is further amplified upon knockdown of p21 in SN cells incubated with Tiron (SN + Tiron/sip21) (Fig. 9G). Importantly, knockdown of p21 resulted in a decrease in HPF-reactive oxidants in SN cells (SN + sip21) (Fig. 9H) but not DHE and mitosox (Fig. 9I and J). On the contrary, a significant increase in HPF and DHE fluorescence, but not mitosox, is detected upon gene specific knockdown of p53 in SN cells (SN + sip53) (Fig. 9H, I & 9J). Lastly, knockdown of p21 does not affect cell cycle arrest in SN cells, however, decreased p53 allows SN cells to start exiting from the S-G2/M boundary (Fig. 9K). The efficiency of gene knockdown of p53 and p21 were confirmed by western blotting (Fig. 9L).

Fig. 9.

Fig. 9

Fig. 9

Open in a new tab

p53, p21 and SA-redox state are part of a regulatory network involved in the establishment of a deep senescent phenotype. A) p21, p53 and γH2AX protein expresssion in pro, SN cells and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC). B) GSH, GSSG and GSH/GSSG ratio assessed in SN cells transfected with a specific p21 siRNA (SN + sip21) relative to SN cells transfected with a control siRNA (SN + siCo) and expressed as fold relative to SN cells transfected with the control siRNA (SN + siCo). Data show are the mean ± SEM of at least three independent experiments. C) IL-6 and IL-8 secretion level in media of pro cells or SN cells transfected with a control siRNA (SN + siCo) or a p21 specific siRNA (SN + sip21) three days after siRNA transfection. Results are expressed as fold relative to pro cells after normalization to cell numbers. The data represent the mean ± SEM of at least three experiments. D) SN + siCo and SN + sip21 cells and E) pro cells transfected with a control siRNA (pro + siCo) or a p21 specific siRNA (pro + sip21) were treated with ABT737 as indicated and cells viability relative to control cells grown in absence of ABT-737 was determined. Data shown are the mean ± SEM of at least three independent experiments. F) Cleaved and full-length caspase 3 in pro + sico, pro + sip21, SN + siCo and SN + sip21 cells incubated with ABT-737 for 48 h. G) pro, SN and SN cells incubated with Tiron (SN + Tiron) transfected with a specific siRNA for p21 or a control siRNA (siCo) were harvested 3 days after the siRNA transfection to assess p21, p53, γH2AX, P–S6 and S6 proteins expression levels. Densitometric analysis of S6 phosphorylation levels was calculated as phosphorylation level relative to total S6 protein level. Phosphorylation level is shown as fold compared to SN + siCo cells. Data represent the mean ± SEM of three independent experiments. H-J) Pro, SN, and SN cells transfected with a control siRNA (SN + siCo), a specific siRNA for p21 (SN + sip21) and a specific siRNA for p53 (SN + sip53) were loaded with H) HPF, I) DHE and J) mitosox fluorescent probes. Bar chart represents fluorescence intensity expressed as percentage of the probes reactivity measured in SN cells. The data represent the mean ± SEM of at least three independent experiments. *p < 0.05 using two-tailed unpaired t-test. K) Cell cycle profiles of pro, SN or SN cells transfected with control siRNA (SN + siCo), a specific p21 siRNA (SN + sip21) and a specific p53 siRNA (SN + sip53). L) Western blot showing the efficacy of SN cells' transfection with control siRNA (siCo), specific p21 siRNA (sip21) or specific p53 siRNA (sip53).

To gain deeper understanding of the relationship between SA-redox state, p21 and p53, we assessed the level of p21 and p53 mRNA in pro, SN, and SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC). While the level of p21 mRNA increases in SN cells compared to pro cells (Fig. 10A), the mRNA of p53 significantly decreased (Fig. 10B). Moreover, while SN cells incubated with Tiron (SN + Tiron) show a significant lower level of p53 mRNA compared to SN cells (Fig. 10B), no difference in p21 mRNA between SN and SN + Tiron cells is observed (Fig. 10A). Intrigued by these data, we set out to evaluate the degradation rate of p21 protein in pro, SN and SN cells incubated with Tiron (SN + Tiron). Results show that p21 protein degrades faster in SN cells compared to pro cells. Interestingly, no significant difference in the degradation rate of p21 is observed between SN and SN cells incubated with Tiron (Fig. 10C and D). On the other hand, the lower degradation rate of p53 observed in SN cells compared to pro cells (Fig. 10 E and F) was even lower in SN + Tiron cells from 16 h of the CHX chase (Fig. 10G and H). Notably, NAC neither affects the expression of p21 or p53 proteins nor the decrease in p53 mRNA in SN cells compared to pro cells.

Fig. 10.

Fig. 10

Open in a new tab

SA-redox state regulates p53 rate of degradation and mRNA level. A) p21mRNA and B) p53 mRNA expression in pro, SN, SN cells incubated with Tiron (SN + Tiron) or NAC (SN + NAC). Data are expressed as fold relative to pro cells and shown as the mean ± SEM of at least three independent experiments. *p < 0.05 comparing SN to pro cells by two-tailed unpaired t-test with Welch's correction. C) Cycloheximide chase in pro, SN and SN + Tiron cells for the indicated time before p21 protein level was assessed. D) Band intensities of p21 were normalized to the loading control β-actin. Results are expressed as fold relative to protein level at “0 h”. Graph shows the mean ± SEM of at least three independent experiments. E) Cycloheximide chase in pro, SN and SN cells incubated with Tiron (SN + Tiron) for indicated time before p53 protein level was assessed. F) Band intensity of p53 was analyzed as in D). G) Cyclohexamide chase in pro, SN and SN + Tiron for 16 h, 24 h and 48 h before p53 protein level was assessed. H) Band intensities of p53 were analyzed as in D).

3. Discussion

In mammals, replicative senescence has evolved to curtail tumorigenesis, but also contributes to organismal aging [42,43]. Hence, while many studies involving cellular senescence use cell models arrested in G1, we chose to induce senescence in S phase synchronized cells to elicit replication stress (RS)-induced senescence and mimic replicative senescence [44]. MMS is known to induce DNA methylation, block elongation of replication forks and trigger replication-associated DNA lesions [45]. As MMS-induced DNA double-strand breaks are most prominent in S-phase of the cell cycle [46,47], upon release from thymidine block, cells re-enter the cell cycle in normal culture medium but encounter RS, thereby resulting in S-G2/M cell cycle arrest and RS-induced senescence. While many studies investigating the relationship between redox stress and cellular senescence are in line with the role of oxidants as DNA damaging agents, we recently proposed that oxidants could also be critical for the establishment of the cellular senescent phenotype [48]. In support of this proposal, we present evidence showing the existence of a O2.--driven redox state termed as SA-redox state involved in the acquisition of key phenotypic features of RS-induced senescence.

3.1. Superoxide is a key oxidant of the SA-redox state

SA-redox state is characterized by increased reactivity with the O2.- sensitive probes DHE, lucigenin, and mitosox as well as the ONOO−/OH. reactive probe HPF, but not the H2O2-reactive probe DCF-DA. The decreased reactivity with the NO-specific probe, DAF, suggests transient consumption of NO due to its reaction with O2.- to generate ONOO−, evidenced by the increase in HPF fluorescence. In contrast to what would be expected with H2O2, SA-redox state mitigates the decrease in GSH without affecting GSH/GSSG ratio. Notably, the glutathione precursor, NAC, commonly used as a scavenger of H2O2, does not appear to regulate SA-redox state; NAC neither affected GSH/GSSG ratio nor the change in reactivity with lucigenin and HPF in the senescent SN cells. In contrast, scavenging intracellular O2.- using Tiron (SN + Tiron) significantly rescued HPF reactivity as well as total GSH levels in SN cells. Taken together, these data support that O2.- is the key determinant of SA-redox state, which involves its reactivity with NO rather than its dismutation to H2O2. These data strongly implicate ONOO− formation as the key event, however, the involvement of OH. radical can not be completely ruled out given its reactivity with HPF [49].

3.2. SA-redox state impacts SASP

An essential hallmark of senescent cells is the acquisition of SASP. The profile of the factors secreted upon the acquisition of SASP is heterogenous, with the exact composition dependent upon the nature of the senescence inducer and the cell type [50,51]. Our data suggest that SA redox state can be an independent regulator of SASP, as seen by the ability of Tiron to significantly reduce the secretion of IL-6 and IL-8 in SN cells. This we show to be an effect of the regulation of NF-κB activity, a master regulator of SASP [33,35]. Alleviating SA-redox state using Tiron inhibited IκBa degradation and p65 nuclear translocation, which is in line with a recent report demonstrating the role of intracellular O2.- and OONO− on the transcriptional activation of NF-κB [52]. However, regulation of NF-κB dependent transcription of SASP factors might not be the only level of regulation that determines the secretory phenotype. This is evidenced by observations delinking regulation at the levels of transcription and translation, such as the case for IL-8, which unlike IL-6, appears to be regulated only at the level of protein secretion. Our data suggest a novel activity of NF-κB, beyond its conventional transcription-linked control of gene expression, whereby inhibiting NF-κB impacted the secretion of IL-8. Whether this is a direct effect or an indirect effect mediated through the expression of a hitherto unknown factor that regulates IL-8 secretion in a redox-dependent manner in SN cells remains to be determined. These data indicate a critical role for SA-redox in determining the profile of SASP factors via activation of NF-κB and are suggestive of an additional level of regulation of the secretory factors.

3.3. SA-redox state, TFEB, S6k and S6 protein phosphorylation and sensitivity to senolysis

SA-redox state induces an increase in TFEB protein expression that was not accompanied by an increase in the transcription of TFEB or a decrease in the rate of the TFEB protein degradation. The exact mechanism involved in redox-dependent increase of TFEB protein needs to be further investigated. It is noticeable that only 64 out of 471 putative direct targets of TFEB described by Palmieri and collaborators are linked to autophagy and lysosome biogenesis [53]. A host of other genes controlled by TFEB are linked to different biological processes like mitochondrial metabolism, regulation of gene expression, cellular response to stress, and cell cycle or translation, among others. TFEB overexpression induces the activation of mitogen-activated protein kinase 1/3 and AKT pro-survival pathways, phosphorylation of mTORC1 effectors 4E-binding protein 1 and S6 kinase B1, and increases protein synthesis [54]. Interestingly, silencing TFEB in SN cells induces a decrease in p21 and S6 protein levels (data not shown), which might imply a novel role for TFEB as a mediator of SA-redox state induced senescence. Sustained activation of the mTOR signaling pathway is another key characteristic of senescent cells, a phenomenon termed as geroconversion. Interestingly, SA-redox state is required to maintain levels of p70S6K and S6 phosphorylation that are comparable to the levels seen in pro cells. As accumulation of geroconverted senescent cells has been shown to influence organismal aging, minimizing geroconversion using redox modulators specifically targeting SA-redox state, such as Tiron, might offer a novel approach to control cellular aging. Along these lines, there is a heightened interest in developing approaches to selectively eliminate senescent cells [5]. Notably, senescent cells show increased sensitivity to drugs, plant extracts, peptides or cell-clearing therapies known as senolytics [1,41]. One such family of molecules is ABT-737, a small-molecule inhibitor of the anti-apoptotic proteins BCL-2, BCL-W and BCL-XL, which triggers targeted elimination of senescent cells [55]. Interestingly, alleviating SA-redox state using Tiron, SN cells transfection with a specific siRNA for p21 or both at the same time (SN + Tiron and SN + sip21, SN + Tiron/sip21) significantly increases sensitivity to senolysis by ABT-737, thus providing a proof of concept approach involving a combination of senolytics and specific redox modulators that target SA-redox state for the efficient execution of senescent cells.

3.4. Interplay between SA-redox state, p53 and p21

p53, p21 and the SA-redox state are shown to be part of a critical regulatory network whereby the SA-redox state increases p21 protein levels, which in return increases the level of HPF reactive oxidants associated with SA-redox state. On the other hand, SA-redox state maintains a low level of p53 expression allowing for an increase in DHE, HPF reactive oxidants. Interestingly, p21-dependent production of oxidants appears to amplify S6 protein phosphorylation and resistance to senolysis. This agrees with the work from Passos et al. demonstrating the presence of a dynamic feedback loop activated by the sustained expression of p21 leading to mitochondrial dysfunction and production of oxidants. This loop was shown to be critical to actively maintain a state of deep cellular senescence. However, Passos et al. only attributed the switch to deep senescence phenotype to the maintenance of DNA damage by p21-dependent ROS production [11]. Our data show that in addition to ensuring active DDR, p21-dependent production of oxidants is key in increasing the level of the SA-redox state oxidants associated with geroconversion as seen by an increase in S6 phosphorylation and resistance to senolysis, two key events associated with cellular aging (Fig. 11).

Fig. 11.

Fig. 11

Open in a new tab

Interplay between SA-redox state, p53 and p21, MMS-induced replication-stress activates DDR. While DDR stabilizes p53 protein a SA-redox state is established. The primary SA-redox state favours the accumulation of p21 protein that in turn amplifies the production of HPF reactive oxidants involved in the maintenance of DDR and mitigation of p53 expression. Primary SA redox state maintains GSH, activates NF-κB, increases the level of TFEB and p21 protein, induces the phosphorylation of S6K and S6 protein. On the other hand, p21-dependent production of oxidants amplifies S6 phosphorylation and provides resistance to senolysis, two key events of the senescent cells geroconversion or deep senescent phenotype associated with cellular aging.

4. Conclusion

We describe the existence of SA-redox state during replication stress-induced senescence that is involved in the activation of key features of the cellular senescent phenotype. In addition, the interplay between p21, p53 and SA-rdox state led us to propose that this p53, p21 and SA-redox state regulatory network might be critical in determing a p53/p21 ratio where a high p53/p21 protein ratio is associated with senescent cells that are responsive to senolysis. On the other hand, a low p53/p21 protein ratio would be associated with geroconverted senescent cells as evidenced by activation of mTOR signaling and resistance to senolytic therapy. From a functional standpoint, O2.- driven redox state described here shares striking similarity with our past work demonstrating the effect of protein modifications, in particular S-nitrosation (S–NO) and 3-nitrotyrosine (3-NT), in cancer cell survival and resistance to execution [[12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32]]. Moreover, the identification of a specific redox milieu and its association with SASP factors present a plausible insight into the functional dichotomy of senescence, whereby the pathological aspects of senescence could be linked to the aberrant redox micro-environment that promotes age-related disorders. Together, our findings unravel a novel signaling node for the specific design and development of therapeutic strategies aimed at selectively targeting SA-redox state in addition to or in lieu of general senolytic approaches. To that effect, in a recent review we argued that the selection of antioxidants with the potential for human health has mostly been done through screens evaluating the ability of the compounds to prevent macromolecular damage and not their effect on cellular senescence. This screening bias might explain why antioxidants supplementation therapy failed to produce the expected outcome in studies involving human disease states associated with aging. Antioxidants, selected based on their capacity to prevent macromolecule damage might not affect the role of oxidants as disruptors of cell signaling involved in the accumulation of senescent cells. Our current data support that an increase in O2.- might be the key event leading to the SA-redox state, hence selecting O2.- specific antioxidants such Tiron or antioxidants that would prevent the formation of HPF reactive oxidants to target specifically the SA-redox might be promising as redox modulators to be used with the intention to not only mitigate DNA damage but also the SA-redox state.

5. Materials and methods

5.1. Cell culture

Human retinal pigmental cells-1 (RPE1-hTERT, ATCC CRL-400) cells were maintained in Dulbecco's Modified Eagle Medium: Nutrient Mixture F-12 (DMEM/F12) with 10% (v/v) fetal bovine serum and 0.01 mg/ml hygromycin B under a humidified environment at 37 °C with 5% CO2. RPE1-hTERT cells were subcultured every four days at around 80% confluency.

5.2. Reagents

Primary antibodies used: 53BP1 (Cat. #SC-22760; Santa Cruz Biotechnology) and LAMP1 (Cat. #ab25630; Abcam). Primary antibodies obtained from Cell Signaling Technology: Ki-67 (Cat. #9129), phospho-histone H2A.X (Ser139) (Cat. #2577), p53 (#9282), FoxM1 (#5436), Phospho-Histone H2A.X (Ser139) (#2577), TFEB (#4240), IκBα (#4812), NF-κB p65 (#8242), Lamin A/C (#4777), Phospho-S6 Ribosomal Protein (Ser235/236) (#2211), S6 Ribosomal Protein (#2217), Phospho-p70 S6 Kinase (Thr389) (#9205), p70 S6 Kinase Antibody (#9202), Cleaved Caspase-3 (Asp175) (#9664), Caspase-3 (#9662). Antibodies for p16 (#554079) and p21 (#556430) were from BD biosciences. Rad51(#ab1837) and Lamin B1 (#ab16048) antibodies were from Abcam. Hsp90 (#SC13119) antibody was from Santa Cruz Biotechnology. β-actin (#A5441) antibody was from Sigma-Aldrich. ABT-737 (#S1002, Selleckchem, Houston, TX), 4,5-dihydroxy-1,3-benzenedisulfonic acid (Tiron) (cat# 172553, Sigma-Aldrich, St. Louis, MO) or 2.5 mM N-Acetyl Cysteine (NAC; Cat# A9165, Sigma-Aldrich).

5.3. Establishment of SN, SN + Tiron, SN + NAC and pro cells

RPE1-hTERT were synchronized in S phase using 4 mM thymidine for 24 h before being exposed to either 4 mM MMS or control culture medium for 1hr. After 1hr, MMS or control medium was replaced by fresh cell culture medium. Senescent phenotype was established 5 days following the initial short exposure to MMS. Cells exposed to MMS are referred to as SN cells and cells exposed to control medium as pro cells. Proliferating pro cells were subcultured when confluent. Fresh medium was added to SN cells twice a week. SN + Tiron and SN + NAC cells were generated by adding Tiron (5 mM) or NAC (2.5 mM), respectively, to SN cells’ culture medium daily from Day 5 to Day 7 after the establishment of day 5 SN cells.

5.4. Cell cycle analysis

SN and pro cells were detached with Accutase (Cat. #00-4555-56; Invitrogen, Thermo Fisher Scientific, Massachusetts, USA), washed with 1xPBS, and fixed with 75% ethanol for 30 min. Cells were then stained with propidium iodide (PI)/RNaseA staining solution for 30 min at 37 °C. PI fluorescence intensity was analyzed using a Becton-Dickinson Fortessa Flow Cytometer (BD Biosciences). Flow cytometry data were analyzed with FlowJo™ Software -BD Bioscience.

5.5. Immunofluorescence detection of Ki-67, LAMP1, γH2AX and 53BP1

Pro and SN cells were grown on coverslips before being fixed with 4% paraformaldehyde for 30 min at room temperature. Cells were permeabilized with 0.2% Triton (or 0.05% saponin for LAMP1) and then incubated with Ki-67, LAMP1, γH2AX and 53BP1 primary antibodies overnight at 4 °C. Images were obtained with Olympus FV1200 Confocal Microscope.

5.6. Senescence-associated β-galactosidase activity

Pro and SN cells were stained for β-galactosidase activity according to manufacturer's recommendation using the Senescence β-Galactosidase Staining Kit (Cat. #9860; Cell Signaling Technology, Massachusetts, USA). Images were captured using the Olympus CKX53 microscope (Tokyo, Japan). Alternatively, the Senescence-associated β-galactosidase activity was measured with the C12FDG protocol as previously described [56].

5.7. IL-6 and IL-8 secretion

Media from pro and SN cells were harvested before the amount of IL-6 and IL-8 secretion was measured using an ELISA based assay following the manufacturer's protocol from R&D systems (Cat. #DY206 and #DY208).

5.8. Detection of oxidants in SN and pro cells

SN and pro cells were detached using Accutase and incubated with 5 μM of the following fluorescent probes from Life Technologies: CM-H2DCFDA (Cat. #C6827), DAF-FM diacetate (Cat. #D23844), Dihydroethidium (Cat. #D11347), HPF (Cat. #H36004), and MitoSOX (Cat. #M36008) for 30 min at 37 °C. Fluorescence intensity was analyzed with a Becton-Dickinson Fortessa Flow Cytometer (BD Biosciences). Flow cytometry data were analyzed with FlowJo™ Software -BD Bioscience. Alternatively, O2.- level was detected using a lucigenin-based chemiluminescence assay. Cells were harvested and permeabilized with 450 μl of somatic cell ATP-releasing reagent (Sigma Aldrich, Missouri, USA). Immediately after cell lysis, 400 μl of the lysate was used for chemiluminescence measurement using the Berthold Sirius Luminometer (Bad Wildbad, Germany). Chemiluminescence was recorded for 30s after 100 μl of 850 mM lucigenin (Sigma Aldrich, Missouri, USA) solution was injected. The intracellular superoxide level is expressed as the luminescence unit after normalization to cell number.

5.9. GSH/GSSG assay

The GSH/GSSG ratio was determined using the measurement of Reduced glutathione (GSH) and oxidized glutathione (GSSG) using the GSH/GSSG-Glo™ Assay (Cat. #V6612; Promega). Experiments were performed following the manufacturer's protocol.

5.10. Nuclear-Cytoplasmic fractionation

Pro, SN, SN + Tiron and SN + NAC cells were detached with Accutase, washed with 1xPBS, and then subjected to cell lysis and stepwise centrifugation for Nuclear-Cytoplasmic fractionation according to manufacturer's recommendation using the NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (Cat. #78835; Thermo Fisher Scientific, Massachusetts, USA). Protein samples from the respective fractions were then analyzed by Western blot.

5.11. Cell viability assay

Cell viability assay was performed on pro and SN cells seeded in 96-well plate. Cell viability assay was performed using the XTT Cell Viability Kit (Cat. #30007; Biotium, California, USA) according to the manufacturer's protocol.

5.12. siRNA-mediated gene knockdown

siRNA-mediated gene knockdown experiments were performed using Lipofectamine™ RNAiMAX Transfection Reagent (Cat. #13778150; Invitrogen, Thermo Fisher Scientific, Massachusetts, USA) according to the manufacturer's protocol. The siRNA for p21 (5′-AGAUUUCUACCACUCCAAAtt-3′; siRNA ID #s417), and control siRNA (Cat. # 4390844) were purchased from Thermo Fisher Scientific (Massachusetts, USA). The siRNA for p53 (L-003329-00) was purchased from Dharmacon (Horizon Discovery).

5.13. Western blot analysis

Cells were lysed using 50 mM Tris pH7.4, 150 mM NaCl, 1 mM EDTA, 0.25% sodium deoxycholate, 1% NP-40, 0.1% SDS lysis buffer supplemented with 1 mM PMSF, 1 μg/ml leupeptin, 1 μg/ml pepstatin A, 1 μg/ml aprotinin, 1 mM sodium fluoride, and 1 mM sodium orthovanadate. Western blot was performed with the same amount of proteins from each sample separated by SDS-PAGE. Protein bands were visualized using Biorad ChemiDoc™ Imaging Systems (Bio-Rad, Hercules, CA, USA). Analysis of Western blot images was performed using Biorad Image Lab™ Software (Bio-Rad, Hercules, CA, USA).

5.14. Real-time PCR

Extraction of RNA from cell pellets were performed using E.Z.N. A® Total RNA Kit I (Cat. #R6834; Omega Bio-TEK, Norcross, GA, USA) according to manufacturer's protocol. A reverse transcription reaction was performed with 1 μg of total RNA using reverse transcription reagents from Applied Biological Materials (Cat. #G490). Real-time quantitative polymerase chain reaction (PCR) reaction was carried out with Precision FAST qPCR Master Mix with ROX and SYBR green (Primerdesign, UK) and quantified with Applied Biosystems 7300 Real-Time PCR System (Thermo Fisher Scientific, Massachusetts, USA). Determination of Ct value was obtained using Auto-Ct function. Gene expression was calculated using the ddCt method with GAPDH mRNA as the endogenous control. The following primers were used for RT-qPCR:

GAPDH:

Forward: ATCTTCCAGGAGCGAGATCC; Reverse: AGAGGGGGCAGAGATGATGA

IL-6:

Forward: GATTCAATGAGGAGACTTGCC; Reverse: TGTTCTGGAGGTACTCTAGGT

IL-8:

Forward: GAGTGGACCACACTGCGCCA; Reverse: TCCACAACCCTCTGCACCCAGT

p21:

Forward: TCTACCACTCCAAACGCC; Reverse: CACAAACTGAGACTAAGGCAG

TFEB:

Forward: CCAGAAGCGAGAGCTCACAGAT; Reverse: TGAGGATGGTGCCCTTGTTC

p53:

Forward: GAGCTGAATGAGGCCTTGGA; Reverse: CTGAGTCAGGCCCTTCTGTCTT

5.15. Statistical analysis

Two-tailed unpaired t-test, two-tailed unpaired t-test with Welch's correction, multiple t tests, one-way ANOVA were performed using GraphPad Prism version 8.0.0 for Windows, GraphPad Software, San Diego, California USA, www.graphpad.com”. p < 0.05 was considered as significant.

Funding

This work was supported by NMRC Singapore Healthy Longevity Catalyst Awards project: Organismal aging and oxidants beyond macromolecules damage (HLCA20Jan-0098) to MVC and Medicine Healthy Longevity Translational Research Program: HLTRP/2021/Rethinking-old-drugs-001 to SP.

Author contributions

Luo Le: Performed the experiments and analyzed data.

SP: Analyzed and edited the final version and partially funded the project.

MVC: Conceptualized and funded the study, analyzed data, wrote the manuscript.

Declaration of competing interest

None of the authors in this study have any conflict of interest to declare.

Acknowledgements

The authors wish to thank Dr. Gireedhar Venkatachalam for his contribution to the initial idea of this project during his doctoral thesis work at National University of Singapore. This work was supported by NMRC Healthy Longevity Catalyst Awards project: Organismal aging and oxidants beyond macromolecules damage (HLCA20Jan-0098) to MVC and Medicine Healthy Longevity Translational Research Program: HLTRP/2021/Rethinking-old-drugs-001 to SP.

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.redox.2023.102757.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Multimedia component 1
mmc1.pdf (194.1KB, pdf)

Data availability

Data will be made available on request.

References

  • 1.Ngoi N.Y., et al. The redox-senescence axis and its therapeutic targeting. Redox Biol. 2021;45 doi: 10.1016/j.redox.2021.102032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Tuttle C.S.L., et al. Cellular senescence and chronological age in various human tissues: a systematic review and meta-analysis. Aging Cell. 2020;19(2) doi: 10.1111/acel.13083. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Wyld L., et al. Senescence and cancer: a review of clinical implications of senescence and senotherapies. Cancers. 2020;12(8) doi: 10.3390/cancers12082134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Zhang L., et al. Cellular senescence: a key therapeutic target in aging and diseases. J. Clin. Invest. 2022;132(15) doi: 10.1172/JCI158450. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Chaib S., Tchkonia T., Kirkland J.L. Cellular senescence and senolytics: the path to the clinic. Nat. Med. 2022;28(8):1556–1568. doi: 10.1038/s41591-022-01923-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Baker D.J., Sedivy J.M. Probing the depths of cellular senescence. J. Cell Biol. 2013:11–13. doi: 10.1083/jcb.201305155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Dulic V., et al. Uncoupling between phenotypic senescence and cell cycle arrest in aging p21-deficient fibroblasts. Mol. Cell Biol. 2000;20(18):6741–6754. doi: 10.1128/mcb.20.18.6741-6754.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Blagosklonny M.V. Geroconversion: irreversible step to cellular senescence. Cell Cycle. 2014;13(23):3628–3635. doi: 10.4161/15384101.2014.985507. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Blagosklonny M.V. Cellular senescence: when growth stimulation meets cell cycle arrest. Aging (Albany NY) 2023;15(4):905–913. doi: 10.18632/aging.204543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Blagosklonny M.V. Rapamycin, proliferation and geroconversion to senescence. Cell Cycle. 2018;17(24):2655–2665. doi: 10.1080/15384101.2018.1554781. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Passos J.F., et al. Feedback between p21 and reactive oxygen production is necessary for cell senescence. Mol. Syst. Biol. 2010;6:347. doi: 10.1038/msb.2010.5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Hirpara J.L., et al. Superoxide induced inhibition of death receptor signaling is mediated via induced expression of apoptosis inhibitory protein cFLIP. Redox Biol. 2020;30 doi: 10.1016/j.redox.2019.101403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Hlaing K.H., Clement M.V. Formation of protein S-nitrosylation by reactive oxygen species. Free Radic. Res. 2014;48(9):996–1010. doi: 10.3109/10715762.2014.942842. [DOI] [PubMed] [Google Scholar]
  • 14.Kumar A.P., et al. Manganese superoxide dismutase is a promising target for enhancing chemosensitivity of basal-like breast carcinoma. Antioxidants Redox Signal. 2014;20(15):2326–2346. doi: 10.1089/ars.2013.5295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Koh L.W., et al. A distinct reactive oxygen species profile confers chemoresistance in glioma-propagating cells and associates with patient survival outcome. Antioxidants Redox Signal. 2013;19(18):2261–2279. doi: 10.1089/ars.2012.4999. [DOI] [PubMed] [Google Scholar]
  • 16.Goo C.K., et al. PTEN/Akt signaling controls mitochondrial respiratory capacity through 4E-BP1. PLoS One. 2012;7(9) doi: 10.1371/journal.pone.0045806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Luo L., Kaur Kumar J., Clement M.V. Redox control of cytosolic Akt phosphorylation in PTEN null cells. Free Radic. Biol. Med. 2012;53(9):1697–1707. doi: 10.1016/j.freeradbiomed.2012.08.566. [DOI] [PubMed] [Google Scholar]
  • 18.Velaithan R., et al. The small GTPase Rac1 is a novel binding partner of Bcl-2 and stabilizes its antiapoptotic activity. Blood. 2011;117(23):6214–6226. doi: 10.1182/blood-2010-08-301283. [DOI] [PubMed] [Google Scholar]
  • 19.Pervaiz S., Clement M.V. Superoxide anion: oncogenic reactive oxygen species? Int. J. Biochem. Cell Biol. 2007;39(7–8):1297–1304. doi: 10.1016/j.biocel.2007.04.007. [DOI] [PubMed] [Google Scholar]
  • 20.Lim S., Clement M.V. Phosphorylation of the survival kinase Akt by superoxide is dependent on an ascorbate-reversible oxidation of PTEN. Free Radic. Biol. Med. 2007;42(8):1178–1192. doi: 10.1016/j.freeradbiomed.2007.01.013. [DOI] [PubMed] [Google Scholar]
  • 21.Akram S., et al. Reactive oxygen species-mediated regulation of the Na+-H+ exchanger 1 gene expression connects intracellular redox status with cells' sensitivity to death triggers. Cell Death Differ. 2006;13(4):628–641. doi: 10.1038/sj.cdd.4401775. [DOI] [PubMed] [Google Scholar]
  • 22.Clement M.V., Sivarajah S., Pervaiz S. Production of intracellular superoxide mediates dithiothreitol-dependent inhibition of apoptotic cell death. Antioxidants Redox Signal. 2005;7(3–4):456–464. doi: 10.1089/ars.2005.7.456. [DOI] [PubMed] [Google Scholar]
  • 23.Pervaiz S., Clement M.V. Tumor intracellular redox status and drug resistance--serendipity or a causal relationship? Curr. Pharmaceut. Des. 2004;10(16):1969–1977. doi: 10.2174/1381612043384411. [DOI] [PubMed] [Google Scholar]
  • 24.Clement M.V., Hirpara J.L., Pervaiz S. Decrease in intracellular superoxide sensitizes Bcl-2-overexpressing tumor cells to receptor and drug-induced apoptosis independent of the mitochondria. Cell Death Differ. 2003;10(11):1273–1285. doi: 10.1038/sj.cdd.4401302. [DOI] [PubMed] [Google Scholar]
  • 25.Pervaiz S., Clement M.V. Hydrogen peroxide-induced apoptosis: oxidative or reductive stress? Methods Enzymol. 2002;352:150–159. doi: 10.1016/s0076-6879(02)52015-2. [DOI] [PubMed] [Google Scholar]
  • 26.Pervaiz S., Clement M.V. A permissive apoptotic environment: function of a decrease in intracellular superoxide anion and cytosolic acidification. Biochem. Biophys. Res. Commun. 2002;290(4):1145–1150. doi: 10.1006/bbrc.2001.6274. [DOI] [PubMed] [Google Scholar]
  • 27.Clement M.V., Pervaiz S. Intracellular superoxide and hydrogen peroxide concentrations: a critical balance that determines survival or death. Redox Rep. 2001;6(4):211–214. doi: 10.1179/135100001101536346. [DOI] [PubMed] [Google Scholar]
  • 28.Pervaiz S., et al. Activation of the RacGTPase inhibits apoptosis in human tumor cells. Oncogene. 2001;20(43):6263–6268. doi: 10.1038/sj.onc.1204840. [DOI] [PubMed] [Google Scholar]
  • 29.Pervaiz S., et al. Superoxide anion inhibits drug-induced tumor cell death. FEBS Lett. 1999;459(3):343–348. doi: 10.1016/s0014-5793(99)01258-2. [DOI] [PubMed] [Google Scholar]
  • 30.Clement M.V., Pervaiz S. Reactive oxygen intermediates regulate cellular response to apoptotic stimuli: an hypothesis. Free Radic. Res. 1999;30(4):247–252. doi: 10.1080/10715769900300271. [DOI] [PubMed] [Google Scholar]
  • 31.Clement M.V., Ponton A., Pervaiz S. Apoptosis induced by hydrogen peroxide is mediated by decreased superoxide anion concentration and reduction of intracellular milieu. FEBS Lett. 1998;440(1–2):13–18. doi: 10.1016/s0014-5793(98)01410-0. [DOI] [PubMed] [Google Scholar]
  • 32.Clement M.V., Stamenkovic I. Superoxide anion is a natural inhibitor of FAS-mediated cell death. EMBO J. 1996;15(2):216–225. [PMC free article] [PubMed] [Google Scholar]
  • 33.Chien Y., et al. Control of the senescence-associated secretory phenotype by NF-kappaB promotes senescence and enhances chemosensitivity. Genes Dev. 2011;25(20):2125–2136. doi: 10.1101/gad.17276711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Zhu Y., et al. Cellular senescence and the senescent secretory phenotype in age-related chronic diseases. Curr. Opin. Clin. Nutr. Metab. Care. 2014;17(4):324–328. doi: 10.1097/MCO.0000000000000065. [DOI] [PubMed] [Google Scholar]
  • 35.Salminen A., Kauppinen A., Kaarniranta K. Emerging role of NF-kappaB signaling in the induction of senescence-associated secretory phenotype (SASP) Cell. Signal. 2012;24(4):835–845. doi: 10.1016/j.cellsig.2011.12.006. [DOI] [PubMed] [Google Scholar]
  • 36.Almowallad S., Alqahtani L.S., Mobashir M. NF-kB in signaling patterns and its temporal dynamics encode/decode human diseases. Life (Basel) 2022;12(12) doi: 10.3390/life12122012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Shin H.M., et al. Inhibitory action of novel aromatic diamine compound on lipopolysaccharide-induced nuclear translocation of NF-kappaB without affecting IkappaB degradation. FEBS Lett. 2004;571(1–3):50–54. doi: 10.1016/j.febslet.2004.06.056. [DOI] [PubMed] [Google Scholar]
  • 38.Adams P.D. Remodeling chromatin for senescence. Aging Cell. 2007;6(4):425–427. doi: 10.1111/j.1474-9726.2007.00313.x. [DOI] [PubMed] [Google Scholar]
  • 39.Franco-Juarez B., et al. TFEB; beyond its role as an autophagy and lysosomes regulator. Cells. 2022;11(19) doi: 10.3390/cells11193153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Weichhart T. mTOR as regulator of lifespan, aging, and cellular senescence: a mini-review. Gerontology. 2018;64(2):127–134. doi: 10.1159/000484629. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Fan Y., et al. Senescent cell depletion through targeting BCL-family proteins and mitochondria. Front. Physiol. 2020;11 doi: 10.3389/fphys.2020.593630. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Campisi J. Aging and cancer: the double-edged sword of replicative senescence. J. Am. Geriatr. Soc. 1997;45(4):482–488. doi: 10.1111/j.1532-5415.1997.tb05175.x. [DOI] [PubMed] [Google Scholar]
  • 43.Warner H.R. Recent progress in understanding the relationships among aging, replicative senescence, cell turnover and cancer. In Vivo. 2002;16(6):393–396. [PubMed] [Google Scholar]
  • 44.Venkatachalam G., Surana U., Clement M.V. Replication stress-induced endogenous DNA damage drives cellular senescence induced by a sub-lethal oxidative stress. Nucleic Acids Res. 2017;45(18):10564–10582. doi: 10.1093/nar/gkx684. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Lundin C., et al. Methyl methanesulfonate (MMS) produces heat-labile DNA damage but no detectable in vivo DNA double-strand breaks. Nucleic Acids Res. 2005;33(12):3799–3811. doi: 10.1093/nar/gki681. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Nikolova T., et al. Homologous recombination protects mammalian cells from replication-associated DNA double-strand breaks arising in response to methyl methanesulfonate. DNA Repair (Amst) 2010;9(10):1050–1063. doi: 10.1016/j.dnarep.2010.07.005. [DOI] [PubMed] [Google Scholar]
  • 47.Pascucci B., et al. The accumulation of MMS-induced single strand breaks in G1 phase is recombinogenic in DNA polymerase beta defective mammalian cells. Nucleic Acids Res. 2005;33(1):280–288. doi: 10.1093/nar/gki168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Clement M.V., Luo L. Organismal aging and oxidants beyond macromolecules damage. Proteomics. 2020;20(5–6) doi: 10.1002/pmic.201800400. [DOI] [PubMed] [Google Scholar]
  • 49.Hogg N., et al. Production of hydroxyl radicals from the simultaneous generation of superoxide and nitric oxide. Biochem. J. 1992;281(Pt 2):419–424. doi: 10.1042/bj2810419. Pt 2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Birch J., Gil J. Senescence and the SASP: many therapeutic avenues. Genes Dev. 2020;34(23–24):1565–1576. doi: 10.1101/gad.343129.120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Basisty N., et al. A proteomic atlas of senescence-associated secretomes for aging biomarker development. PLoS Biol. 2020;18(1) doi: 10.1371/journal.pbio.3000599. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Yee Y.H., et al. Sustained IKKbeta phosphorylation and NF-kappaB activation by superoxide-induced peroxynitrite-mediated nitrotyrosine modification of B56gamma3 and PP2A inactivation. Redox Biol. 2021;41 doi: 10.1016/j.redox.2020.101834. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Palmieri M., et al. mTORC1-independent TFEB activation via Akt inhibition promotes cellular clearance in neurodegenerative storage diseases. Nat. Commun. 2017;8 doi: 10.1038/ncomms14338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Torra A., et al. Overexpression of TFEB drives a pleiotropic neurotrophic effect and prevents Parkinson's disease-related neurodegeneration. Mol. Ther. 2018;26(6):1552–1567. doi: 10.1016/j.ymthe.2018.02.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Yosef R., et al. Directed elimination of senescent cells by inhibition of BCL-W and BCL-XL. Nat. Commun. 2016;7 doi: 10.1038/ncomms11190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Debacq-Chainiaux F., et al. Protocols to detect senescence-associated beta-galactosidase (SA-betagal) activity, a biomarker of senescent cells in culture and in vivo. Nat. Protoc. 2009;4(12):1798–1806. doi: 10.1038/nprot.2009.191. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1
mmc1.pdf (194.1KB, pdf)

Data Availability Statement

Data will be made available on request.

Articles from Redox Biology are provided here courtesy of Elsevier

RESOURCES