High Hemin Concentration Induces Escape from Senescence of Normoxic and Hypoxic Colon Cancer Cells

Simple Summary High red-meat consumption as well as bleeding or bruising can promote oxidative stress and, in consequence, cancer development. However, the mechanism of that phenomenon is not understood. The induction of therapy-induced senescence (TIS) might also be induced by oxidative stress. Recently, TIS cells, despite their inhibited proliferation potential, have been identified as one of the sources of tumor re-growth. Here, with the use of molecular analyses, we found that oxidative stress, promoted by high doses of hemin or H2O2, can trigger TIS escape and cell re-population. It is closely related to the activity of antioxidative enzymes, especially heme oxygenase-1. Hypoxia might accelerate these effects. Therefore, we propose that the prevention of excessive oxidative stress could be a potential target in senolytic therapies. Abstract Hemoglobin from either red meat or bowel bleeding may promote oxidative stress and increase the risk of colorectal cancer (CRC). Additionally, solid cancers or their metastases may be present with localized bruising. Escape from therapy-induced senescence (TIS) might be one of the mechanisms of tumor re-growth. Therefore, we sought to study whether hemin can cause escape from TIS in CRC. To induce senescence, human colon cancer cells were exposed to a chemotherapeutic agent irinotecan (IRINO). Cells treated with IRINO exhibited common hallmarks of TIS. To mimic bleeding, colon cancer cells were additionally treated with hemin. High hemin concentration activated heme oxygenase-1 (HO-1), induced escape from TIS and epithelial-to-mesenchymal transition, and augmented progeny production. The effect was even stronger in hypoxic conditions. Similar results were obtained when TIS cells were treated with another prooxidant agent, H2O2. Silencing of antioxidative enzymes such as catalase (CAT) or glutathione peroxidase-1 (GPx-1) maintained colon cancer cells in a senescent state. Our study demonstrates that a high hemin concentration combined with an increased activity of antioxidative enzymes, especially HO-1, leads to escape from the senescence of colon cancer cells. Therefore, our observations could be used in targeted anti-cancer therapy.


Introduction
Colorectal cancer is a major public health problem, being the third most commonly diagnosed cancer and the fourth cause of cancer-related death worldwide [1]. Epidemiologic factors related to an increased risk for CRC include old age, tobacco use, red meat species (ROS) [25], here, we checked, whether antioxidative systems might interfere with the induction of senescence/escape from senescence. To induce senescence, HCT116 and SW480 colon cancer cells were treated with irinotecan (IRINO) and a series of hemin concentrations for 24 h. After, that medium was changed to drug-free one and cells were cultured for the next 4 days ( Figure 1A). The induction of senescence was correlated with differential changes in the expression of antioxidative enzymes. The expression of CAT and GPx-1, enzymes involved in the reaction to H 2 O 2 treatment, was elevated ( Figures 1B and A1A,B). In contrast, the expression of heme oxygenase-1 (HO-1), an enzyme that converts heme to biliverdin, CO, and iron ions, was reduced ( Figures 1B and A1C). To verify the effects of HO-1 on the induction of senescence, colon cancer cells were treated with hemin, as an HO-1 activator and substrate. In senescent cancer cells, hemin induced HO-1, CAT, and GPx-1 expression ( Figures 1B and A1A-C). Hemin, only at the highest dose of 100 µM, showed a strong tendency to decrease the metabolic activity of cancer cells ( Figure 1C, p = 0.054; Figure A10A). It also showed the tendency to increase the total cell number ( Figure 1D) but not the proportion of cells in the subG1 phase ( Figures 1E and A10B). These data suggest that hemin at high concentrations may affect the senescence process.
Hemin at 100 µM concentration tends to reduce the proportion of senescent cells, as shown by decreased numbers of SA-β-gal-positive cells (Figure 2A,B, p = 0.052; Figure A10C,D). It correlated with higher proportions of cells in the G0/G1 phase ( Figures 2C and A10E). Recently, we showed that such an alteration in cell cycle distribution might be related to amitotic divisions of senescent cancer cells and the appearance of a small, highly proliferating progeny [27]. In contrast, the proportion of cells accumulated in the G2/M phase was reduced ( Figures 2D and A10F). There were no changes in the proportion of senescent HCT116 cells accumulated in the S ( Figure A11A) or polypoidal phases upon hemin treatment ( Figure A11B). In the case of SW480 cells, a tendency toward a reduction in cells present in the S ( Figure A11C, p = 0.056) and polypoidal phases was observed ( Figure A11D). To confirm that a high concentration of hemin induced escape from senescence, the expression of proteins involved in proliferation and senescence was checked. The levels of proteins related to the cell cycle progression-cyclin A, cyclin B, p-cdc2, and p-Rb-were downregulated in IRINO-treated cells but upregulated when senescent cells were treated with 100 µM of hemin (Figures 2E and A1D-G). In this vein, the expression of two proteins related to cell cycle inhibition and geroconversion-p21 and p-S6-was reduced ( Figures 2F and A2A,B). The expression of PARP-1, a protein responsible for DNA repair, was upregulated in hemin-treated senescent cancer cells ( Figures 2F and A2C), whereas the expression of γ-H2AX, the sensor of DNA double breaks, showed a tendency to reduce ( Figures 2F and A2D). Accordingly, the presence of the cleaved form of PARP-1, indicative of the ongoing apoptosis process, was not detected ( Figures 2F and A2C).
Altogether, these data showed that, in contrast to CAT and GPx-1, HO-1 expression was downregulated upon the induction of senescence in colon cancer cells. A high concentration of hemin, the HO-1 substrate, and the activator most likely promoted escape from senescence. It was marked by the appearance of SA-β-gal negative, highly proliferating cells, progressing through the G0/G1 phase, and a shift from senescence to proliferation molecular markers.     As high hemin concentration results in increased oxidative stress [37,38], we tested whether the prooxidant, H 2 O 2 , would lead to escape from senescence, too. HCT116 cells were incubated with 5 or 10 µM H 2 O 2 and IRINO for 24 h and then cultured in a drug-free medium for the next 4 ("1 + 4" protocol) or 7 ("1 + 7" protocol) days ( Figure 3A). H 2 O 2 at the 10 µM concentration caused a significant decrease in cell numbers at day 1 + 4 ( Figure 3B). It correlated with elevated numbers of SA-β-gal positive ( Figure 3C,D) and granular cells ( Figure 3E). At day 1 + 7, upon senescence escape, the percentages of SA-β-gal and granular cells were significantly reduced ( Figure 3C-E).
Moreover, on day 1 + 4, H 2 O 2 -treated senescent cells showed reduced proportion in the G0/G1 phase of the cell cycle ( Figure 4B) and an augmented accumulation in the G2/M phase ( Figure 4D). On day 1 + 7, upon senescence escape, the effects were opposites ( Figure 4B,D). There was also a significant decrease in the proportion of H 2 O 2 -treated cells accumulated in the subG1 phase between days 1 + 7 and 1 + 4 ( Figure 4A). Escape from senescence correlated with an increased proportion of cells in the S phase, which was the most visible in cells treated with the highest H 2 O 2 dose ( Figure 4C). There were no changes in the proportion of polyploid cells ( Figure 4E). These data suggest that increased oxidative stress related to high H 2 O 2 dose promotes an escape from senescence.
To verify these observations, the expression of proteins related to senescence and proliferation was tested. On day 1 + 4, the expression of a senescent marker, p21 ( Figures 5A and A3A), was elevated, whereas the expression of the proliferation markerscyclin A (tendency), cyclin B, p-cdc2 (tendency), and p-Rb-was downregulated (Figures 5B and A3D-G). On day 1 + 7, upon senescence escape, the effect was the opposite ( Figures 5A,B and A3A,D-G). There were no changes in p53 and p-S6 expression ( Figures 5A and A3B,C). HO-1 expression was elevated upon escaping from senescence. The effect was even stronger when senescent cells were treated with H 2 O 2 ( Figures 5C and A4C). In contrast, upon senescence escape, H 2 O 2 -treated cells exhibited a reduced expression of CAT ( Figures 5C and A4A), whereas GPx-1 expression did not change ( Figures 5C and A4B). As we showed previously, escaping from senescence and progeny production might be related to acquiring a stem cell phenotype and epithelial-to-mesenchymal transition (EMT) [26,27]. In line with these data, senescent HCT116 cells treated with H 2 O 2 underwent mesenchymal-to-epithelial transition (MET): at day 1 + 4, they exhibited augmented expression of E-cadherin ( Figures 5D and A4D) and reduced expression of its inhibitor, Snail ( Figures 5D and A4E). At day 1 + 7, when cells started to proliferate, the opposite process (EMT) was observed ( Figures 5D and A4D,E). There were no changes in the expression of a stem cell marker, Nanog ( Figures 5D and A4F).
To summarize, these data suggest that increased oxidative stress due to a high H 2 O 2 concentration, similarly to high hemin concentration, triggers an escape from senescence, the appearance of highly proliferating progeny, a shift in HO-1/CAT expressions, and EMT. High H2O2 leads to escape from senescence at day 1 + 7: increased cell number but decreased SA-β-Gal and granularity. (A) The scheme of the experiment. Cells were subjected to 2.5 µM of irinotecan and 5 or 10 µM H2O2. After 24 h, the medium was changed and the cells were cultured in a drug-free medium for 4 (■) or 7 (□) days. In a 7-day long variant of the experiment, Figure 3. High H 2 O 2 leads to escape from senescence at day 1 + 7: increased cell number but decreased SA-β-Gal and granularity. (A) The scheme of the experiment. Cells were subjected to 2.5 µM of irinotecan and 5 or 10 µM H 2 O 2 . After 24 h, the medium was changed and the cells were cultured in a drug-free medium for 4 ( ) or 7 ( ) days. In a 7-day long variant of the experiment, after 4 days of cell culture, the medium was changed to the new one. (B) Evaluation of cell number on the 4th and 7th days after the treatment. (C) Activity of SA-β-gal enzyme in UNT and IRINO-treated cells after treatment with H 2 O 2 . Cells were cytospined, and cytochemical staining for SA-β-gal activity was performed. Representative photos were acquired using light microscopy. Original magnification

Silencing of Antioxidative Enzymes Maintains Colon Cancer Cells in a Senescent State
To assess the possible molecular mechanisms responsible for the observed changes, HO-1, CAT, and GPx-1 were silenced with specific siRNAs. After gene silencing, colon cancer cells were treated with IRINO and a 100 µM concentration of hemin. GPx-1 knockout slightly decreased the total cell number ( Figure 6A) and increased the numbers of SA-β-gal positive cells in the presence of hemin ( Figure 6B,C). Hemin also reduced the subpopulation of SA-β-gal positive cells among CAT-silenced ones ( Figure 6B,C). HO-1 silencing did not affect cell number ( Figure 6A) or the proportion of SA-β-gal positive cells ( Figure 6B,C). cancer cells were treated with IRINO and a 100 µM concentration of hemin. GPx-1 knockout slightly decreased the total cell number ( Figure 6A) and increased the numbers of SAβ-gal positive cells in the presence of hemin ( Figure 6B,C). Hemin also reduced the subpopulation of SA-β-gal positive cells among CAT-silenced ones ( Figure 6B,C). HO-1 silencing did not affect cell number ( Figure 6A) or the proportion of SA-β-gal positive cells ( Figure 6B,C). In HCT116 cells, genes encoding antioxidative enzymes-HO-1, CAT, or GPx-1-were silenced with the use of siRNAs. Cells were exposed to IRINO Figure 6. Silencing of antioxidative enzymes maintains colon cancer cells in a senescent state: changed proportion of SA-β-Gal positive cells. In HCT116 cells, genes encoding antioxidative enzymes-HO-1, CAT, or GPx-1-were silenced with the use of siRNAs. Cells were exposed to IRINO and 100 µM hemin ( ). Control cells were subjected to IRINO and NaOH treatment ( ). Cells were incubated with drugs for 24 h and then cultured in a drug-free medium for 4 days. (A) Evaluation of cell number. Cell number was counted with the use of Bürker's chamber; (B) Evaluation of SA-β-gal positive cells. (C) Activity of SA-β-gal enzyme. Cells were cytospined and cytochemical stained for SA-β-gal activity. Representative photos were acquired using light microscopy. Original magnification 200×, scale bar 50 µM; Each bar represents mean ± SEM, n ≥ 3; ## p < 0.01-vs. si-NEG hemin 100 µM; $$ p < 0.01-control vs. hemin 100 µM.
Then, the expression of proteins related to oxidative stress, cell proliferation, senescence, EMT, and stemness was checked. Firstly, successful silencing of CAT and GPx-1 was confirmed ( Figures 7A and A5A,B). In the case of HO-1, gene knock-out was not entirely effective and the expression of HO-1 was further induced in response to hemin ( Figures 7A and A5C). This can explain why no significant effects of HO-1 silencing on cell number ( Figure 6A) or proportion of SA-β-gal positive cells were observed ( Figure 6B,C). Nevertheless, HO-1 silencing led to a reduced expression of proliferation marker p-Rb, regardless of hemin presence or absence ( Figures 7B and A5G). It was accompanied by elevated levels of the DNA repair protein PARP-1 ( Figures 7C and A6C) and the mesenchymal marker Snail (Figures 7D and A6E). Upon hemin addition, Snail expression was downregulated ( Figures 7D and A6E), whereas the level of the stem cell marker Nanog was augmented ( Figures 7D and A6F). The silencing of CAT or GPx-1 showed a tendency to decrease the expression of proliferation marker cyclin B in the presence of hemin ( Figures 7B and A5E, p = 0.057). A loss of GPx-1 also reduced the expression of p-Rb, another proliferation marker, both in the presence and absence of hemin ( Figures 7B and A5G), while PARP-1 expression was augmented ( Figures 7C and A6C). It also caused MET, as an expression of E-cadherin, to be upregulated, whereas the expression of Snail was downregulated ( Figures 7D and A6D,E). In the presence of hemin, it seemed that the opposite process of EMT was started, as E-cadherin expression was reduced ( Figures 7D and A6D). It was correlated with a decreased expression of the cell cycle inhibitor p21 ( Figures 7C and A6A). Snail expression was decreased in both CAT-and GPx-1-silenced cells treated with hemin ( Figures 7D and A6E). Silencing of antioxidative enzymes maintains colon cancer cells in a senescent state: changed expression of proteins related to proliferation and EMT. In HCT116 cells, genes encoding antioxidative enzymes-HO-1, CAT, and GPx-1-were silenced with the use of siRNAs. Cells were exposed to IRINO and 100 µM hemin ( ). Control cells were subjected to IRINO and NaOH treatment ( ). Cells were incubated with drugs for 24 h, then cultured in a drug-free medium for 4 days. Taken together, these data suggest that the silencing of antioxidative enzymes-HO-1, CAT, and GPx-1-maintains colon cancer cells in a non-proliferating, epithelial state with increased mechanisms of DNA protection. Upon hemin addition, cells started to show features of cancer stem cells and escape from senescence.

Hemin Enhances Escape from Senescence in Hypoxic Cells in the Presence of CAT
As shown previously in colon and lung cancer cells, hypoxia may significantly affect the response of cancer cells to chemotherapeutics, including the development of senescence/escaping from it [39]. Therefore, the role of hypoxia on senescence escaping, upon modulation of oxidative status, was tested here.
Low oxygen tension significantly decreased the total cell number ( Figure 8A) and the proportion of SA-β-gal positive cells in the HCT116 cell line ( Figure 8B,C). In SW480 cells, hypoxia also decreased total cell number, whereas hemin application increase its rate ( Figure A12A). However, hypoxia did not change the proportion of SA-β-gal positive cells in SW480 cells ( Figure A12B,C). We previously showed that escaping from senescence is a multistep process that depends on the cell line, drug type and its concentration, the timing of the experiment, etc. We proposed that it can be monitored using different cellular and molecular markers [26,27,39]. In this vein, SW480 cells demonstrated a reduced accumulation in the subG1 ( Figure A13A) and S phases ( Figure A13D), as well as progression from the G2/M to G0/G1 phases upon hemin treatment ( Figure A13B,C). There were no changes in the polyploid phase ( Figure A13E). The effects were more pronounced in hypoxia ( Figure A13 Figures 9A and A7C). Hemin more strongly induced HO-1 in hypoxia than in normoxia, regardless of CAT status ( Figures 9A and A8C). In contrast, it did not affect CAT and GPx-1 expression in hypoxia differently than in normoxia ( Figures 9A and A8A,B). In terms of proliferation markers, only two proteins-cyclin A and cyclin B-were affected, and their expression seemed to be balanced by hypoxia, hemin, and CAT expression ( Figures 9B and A8D,E). The expression of cyclin A was upregulated in CAT-silenced cells in normoxia, whereas in hypoxia, the effect was the opposite ( Figures 9B and A8D). In contrast, the expression of cyclin B was elevated in CAT-silenced cells, but only in hypoxia. Hemin seemed to compensate for these effects ( Figures 9B and A8E). In this vein, hemin downregulated the expression of cell cycle inhibitor p21 regardless of oxygen tension and CAT expression ( Figures 9C and A9A). A similar effect was observed for another cell cycle inhibitor, p53, but only in normoxia and in the presence of CAT ( Figures 9C and A9C). Hemin also reduced the epithelial marker expression for E-cadherin both in normoxia and hypoxia. The observed effects were reversed in the absence of CAT ( Figures 9D and A9E). There were no changes in the expression of p-cdc2 ( Figures 9B and A8F Altogether, these data showed that hypoxia reduced the expression of proteins related to proliferation, cell cycle inhibition, and antioxidative defense, whereas those related to EMT/MET and DNA repair were not changed. Upon hemin addition, there was HO-1 and cyclins activation, followed by p21 reduction and EMT. CAT expression maintained a reduced expression of E-cadherin and further upregulated cyclins' expression.

Figure 8.
Hemin enhances escaping from senescence in hypoxic cells. In HCT116 cells, the gene encoding CAT was silenced with the use of siRNAs. Cells were cultured in parallel in normoxia or hypoxia. Cells were exposed to 100 µM hemin (□). Control cells were subjected to IRINO and NaOH in the corresponding dose (■). (A) Evaluation of cell number. Cell number was counted with the use of Bürker's chamber. (B) Evaluation of SA-β-gal positive cells; (C) Activity of SA-β-gal enzyme. Cells were cytospined and cytochemical staining for SA-β-gal activity was performed. Representative photos were acquired using light microscopy. Original magnification 200×, scale bar 50µM; Each bar represents mean ± SEM, n ≥ 3; ** p < 0.01, *** p < 0.001-vs. si-NEG control normoxia; $ p < 0.05control vs. hemin 100 µM.  Representative blots show levels of E-cadherin, Snail, and Nanog. GAPDH or vinculin acts as a loading control.
Hemoglobin from either red meat or bowel bleeding may serve as an enhancer of oxidative damage in the bowel and increase the risk of colorectal cancer [4,5]. However, molecular mechanisms beyond that phenomenon remain unclear. Therefore, here, we asked the question of whether the oxidative stress induced by hemin can play a role in the escape of cancer cells from senescence. The most important findings of our study are the following: (1) high concentration of prooxidants, hemin, and H 2 O 2 can trigger escapes from senescence of colon cancer cells, which is manifested by the appearance of a highly proliferating progeny and a shift in EMT/MET markers; (2) silencing of antioxidative enzymes, CAT, GPx-1, and HO-1 maintain senescent cells in the non-proliferating state; (3) in contrast to CAT and GPx-1, the expression of HO-1 is reduced upon the induction of senescence but is restored when cancer cells escape from it; and (4) hemin enhances the escape from senescence in hypoxic cells through HO-1/cyclins activation.
In our previous studies, we demonstrated that human colon cancer cells treated with chemotherapeutic agents-doxorubicin, irinotecan, and 5-fluorouracil-underwent senescence in long-term cell cultures, but after withdrawal of the drug, cell re-growth was observed [25][26][27]36]. Moreover, a subpopulation of TIS cells exhibited certain features of cancer stem cells, namely, elevated Nanog expression, augmented proportion of CD24 + cells and side population, epithelial-to-mesenchymal transition (EMT), and tumor formation in NOD/SCID animals [26,27]. Additionally, in another study, we showed that doxorubicin-treated HCT116 cells displayed an increased production of reactive oxygen species (ROS), whereas a decreased level of ROS prevented the cells from escaping senescence [25]. Therefore, here, we asked the question of whether antioxidative systems can play a role in cancer cells escaping from senescence. In the present study, we showed that the induction of senescence in colon cancer cells led to the upregulation of two enzymes involved in H 2 O 2 utilization, catalase, and glutathione peroxidase-1 but the downregulation of heme oxygenase-1, heme degrading enzyme. To check whether we would be able to reverse this effect, we treated HCT116 and SW480 cells with hemin, a major substrate and an activator of HO-1. In accordance, we demonstrated that at a high concentration (100 µM) hemin led to HO-1 induction and escape from senescence. It correlated with the appearance of SA-β-galactosidase negative cells, progression through the G0/G1 phase, and upregulation of proliferation and EMT markers. Similar effects were obtained when senescent cells were treated with another prooxidant, H 2 O 2. The overexpression of HO-1 has been demonstrated in many cancers [55][56][57] and its level could be further increased in response to oxidative stress, chemo-, radio- [58], or photodynamic therapy [59][60][61]. Moreover, we showed that HO-1 may exert potent and complex effects depending on the cancer type and microenvironmental context [59,[62][63][64][65][66][67][68][69][70]. However, in the perspective of senescence and escaping from senescence of cancer cells, the role of HO-1 has not been studied yet. Interestingly, in a model of myeloid cells, Hedblom and coworkers demonstrated that the deletion of HO-1 resulted in an impaired DNA damage response (DDR), reduced cell proliferation, and increased cellular senescence [71]. Additionally, there are few reports of solid cancers or their metastases presenting with localized bruising. Bruising has been described in parathyroid adenomas [72], orbital neuroblastoma [73], neurofibromatosis [74], and malignant melanoma [75,76]. These data may suggest that the heme/HO-1 system may participate in tumor initiation and metastases. In accordance, the role of HO-1 in cell dedifferentiation/stemness has been shown in several models [66]. For example, Kim and coworkers demonstrated that HO-1-derived CO production participated in the manifestation of breast cancer stem cell-like properties and stimulated the formation of mammospheres through activation of Notch-1 signaling [77]. In our recent paper, we showed that HO-1 had effects on melanoma stem cells' properties [69]. We and others also showed that TIS cells may exhibit certain features of cancer-initiating cells in the dormant state [22,26,27,50,54,78,79]. Polyploid, giant, cancer cells (PGCCs) containing an elevated genomic content and resembling blastocyst might be key players in the evolution of cancer and long-term resistance to therapies [27,34,35]. Additionally, we demonstrated that hypoxia increased the escape from senescence of lung and colon cancer cells treated with chemotherapeutic drugs [39]. In this vein, Zhang and coworkers reported that PGCCs, which possess features of cancer-initiating cells, could be induced from ovarian cancer cells by the hypoxia-mimetic agent, CoCl 2 [54]. They showed that PGCCs from ovarian cancer could grow into tumor spheroids, initiate tumor growth in nude mice, and differentiate into other benign cell types in vitro and in vivo [54,[79][80][81]. These properties of PGCCs were also reported in colon cancer [82]. Interestingly, Li and coworkers demonstrated that PGCCs also produced embryonic hemoglobin-delta and -zeta with a strong oxygen-binding ability and erythroid differentiation-related proteins, promoting the survival of tumor cells in a hypoxic microenvironment [83]. Here, we showed that hemin enhances escape from senescence in hypoxic colon cancer cells through HO-1/cyclins activation. In line with our data, exposure to hemin increased the gene and protein expression level of HO-1 and diminished the ROS levels, senescence, and the inflammatory profile in fibroblasts isolated from the lung biopsies of patients with chronic obstructive pulmonary disease. It also rescued mitochondria dysfunction by restoring mitophagy [84]. Moreover, HO-1 improved heart function and attenuated cardiomyocyte senescence triggered by ischemic injury and aging [85], whereas hemin enhanced the cardioprotective effects of mesenchymal stem cell-derived exosomes against infarction via amelioration of cardiomyocyte senescence [86]. Interestingly, Luo and coworkers showed that HO-1 was accumulated in the nuclei of stress-induced senescent endothelial cells and conferred protection against endothelial senescence independent of its enzymatic activity [87]. Altogether, our data suggest that the hemin/HO-1 system may participate in escaping from the senescence of cancer cells. Therefore, we propose that it could be used in targeted anti-cancer therapy.
In this study, we also showed that, similarly to hemin, another prooxidant H 2 O 2 leads to an escape from senescence. Hydrogen peroxide is freely diffusible and is a relatively longlived molecule. Although it is not very reactive, it is the precursor of many other reactive oxygen species (ROS). One of the most important products is a strongly reactive hydroxyl radical (·OH). The reaction takes place in the presence of transition metal ions such as Fe 2+ using the Fenton reaction. Among the by-products of heme degradation by HO-1 are iron ions. H 2 O 2 , ·OH, and other ROS are essential for various biological processes, e.g., proliferation or differentiation in normal and cancer cells [88]. ROS has been shown to play tumor-promoting but also tumor-suppressing actions [89]. Importantly, ROS production is a mechanism shared by most chemotherapeutics [90]. Recent evidence suggests that prolonged chemotherapy can reduce the overall cellular ROS in cancer, which is believed to function as a key underlying mechanism of drug resistance in chemotherapy [91]. SIPS might be induced by various stressful agents, including ROS or chemotherapeutics. Some data suggest that ROS may play a role not only in the induction of senescence but also in its escape. Gosselin and collaborators reported that oxidative stress induced epithelial cell senescence and that such cells were able to divide by budding [92]. Similarly, Achuthan and coworkers demonstrated that a senescent population of cancer cells had a high level of ROS, whereas their progeny displayed stem-like cell features but low ROS production [93].
In accordance, our previous study showed that the ROS scavenger, Trolox, protected from the emergence of escapers [25]. Recently, Dharmalingam and coworkers showed that, while free iron produced mostly reactive oxygen species (ROS)-related single-strand DNA breaks, hemin induced rapid and persistent nuclear and mitochondrial double-strand breaks (DSBs) in neuronal and endothelial cell genomes and mouse brains following experimental intracerebral hemorrhage. These were comparable to that seen with γ-radiation and DNAcomplexing chemotherapies. While they applied antioxidant therapy to prevent senescence, cells became again sensitized to ferroptosis [94]. Here, we demonstrated that, upon the induction of senescence with IRINO, the expression of CAT and GPx-1, enzymes that utilize H 2 O 2 , was strongly upregulated in colon cancer cells. Moreover, when cells were treated with high hemin concentration, which led to escaping from senescence, the CAT and GPx-1 levels were further elevated. In turn, when senescent cells were treated with H 2 O 2 , the CAT expression was maintained at a high level during the whole experiment (7 days), dropping down by the end of it. It correlated with the strong increase in proteins related to cell cycle progression-cyclin A, cyclin B, p-cdc2, and p-Rb-and the appearance of a small, highly proliferating progeny. Interestingly, when we silenced CAT and GPx-1 in senescent colon cancer cells with specific siRNAs and treated these cells with hemin, we observed that the expressions of cyclins, especially cyclin B, were reduced. As cyclin B is necessary for the progression of the cells into and out of the M phase of the cell cycle, we propose that excessive ROS production, unbalanced by activities of antioxidative enzymes CAT and GPx-1, inhibits escaping from senescence. The intrinsic levels of antioxidant enzymes are low in a majority of cancer cell types as compared to non-transformed cells. This suggests that those cancer cells may lack the biochemical machinery, which is crucial for detoxifying higher fluxes of H 2 O 2 . Kinetic models built using in vitro data have shown that CAT is the major enzyme responsible for the detoxification of high concentrations of H 2 O 2 , whereas GPx-1 is involved in removing low fluxes of H 2 O 2 [95]. Our data suggest that upregulation of CAT and GPx-1 in senescent cancer cells may be a characteristic feature and may help distinguish them from non-senescent cancer cells. Additionally, we propose that this phenomenon could be responsible for the enhanced resistance of senescent cancer cells to oxidative stress and oxidative stress-based anticancer therapies.
Taken together, this work shows that high hemin or H 2 O 2 concentrations could be a trigger for colon cancer cells to escape from senescence. However, the outcome depends on the proper activities of antioxidative enzymes: HO-1, CAT, and GPx-1. HO-1 expression, which is reduced upon induction of senescence, seems to be restored when cancer cells escape from it. Oppositely, the CAT and GPx-1 levels are strongly elevated in cancer cells that undergo TIS. Hemin enhances escaping from senescence in hypoxic colon cancer cells through HO-1/cyclins activation. Therefore, our data indicate that the oxidative status of senescent cancer cells differs from their non-senescent counterparts and we suggest its use in senolytic-oriented therapy. A thorough combination of prooxidants and/or antioxidative enzymes' activity could be a valuable help to prevent senescent cells from re-activating their proliferative properties and cancer re-growth after chemotherapies.

Chemicals and Antibodies
Unless otherwise specified, chemicals and reagents were purchased from Sigma Aldrich (Taufkirchen, Germany). A list of the antibodies is presented in the table below (Table 1):
To induce senescence, cells were seeded at a density of 250 000/per 25 cm 2 flask 24 h before treatment. Colon cancer cells were cultured in the presence of 2.5 µM (HCT116) or 2.5 µM (SW480) irinotecan (IRINO) for 24 h followed by 4-7 days in a fresh medium without a drug. In the experiments, where growing time lasted for 7 days after the incubation with the drugs, the medium was changed to the fresh one on day 4. Doses of IRINO used in experiments were chosen as most efficient to induce senescence but not cell death, based on staining for SA-β-gal activity, cell counting in Bürker's chamber, BrdU incorporation assay, and cell cycle analysis with PI staining [27]. Additionally, cells were incubated with hemin or H 2 O 2 during the chemotherapeutic treatment. Incubation with drugs lasted 24 h; after that time, the medium was changed to drug-free one and the cells were grown for additional 4 or 7 days. NaOH was used as a solvent for hemin. Cells treated with NaOH were used as a control in the hemin-based experiments. The experiments were performed in technical duplicates or triplicates. All tests were performed in at least three biological repetitions.

Experiments in Hypoxia
In the experiments performed on 96-well plates, cells were seeded at 1500 cells/per well for HCT116 or SW480 cells. For the remaining tests, 125,000 cells were plated on a 25 cm 2 flask for each cell line. After seeding, cells were grown in normoxia for 24 h. Half of the plates/flasks were then left in normoxic condition, whereas the remaining culture dishes were transferred to the hypoxic chamber, where the medium was changed to the hypoxic one. The hypoxic medium was incubated for at least 24 h in the hypoxic chamber before the experimental procedure. After 24 h of culture in normoxia or hypoxia, cells were treated with chemotherapeutics for the next 24 h. After 24 h of incubation with chemotherapeutic drugs, the medium was changed and the cells were cultured in a drug-free medium for the next 7 days. Additionally, the medium was changed on day 4.

Gene Silencing
HCT116 cells were seeded at 750 per well or 62,500 cells per 25 cm 2 flask in normoxic conditions. Twenty-four hours later, half of the plates/flasks were left in normoxic condition, whereas the remaining culture dishes were transferred to the hypoxic chamber and the medium was changed. After a 24 h incubation, the cells were transfected with pooled siRNAs targeting: HO-1, GPx-1, or CAT (ON-TARGETplus siRNA, Dharmacon, Lafayette, CO, USA) using lipofectamine RNAiMAX (Invitrogen, Life Technologies Corporation, Carlsbad, CA, USA). Nontargeting siRNAs were used as a control. Lipofectamine and siRNA dilutions were prepared according to manufacturers' protocols. The next day, cancer cells were treated with respective chemotherapeutics for 24 h followed by 4 days of drug-free culture. Afterward, the effects of gene silencing on protein expression, cell number, and SA-β-gal activity were analyzed. Optical density was measured at 562 nm using a scanning multi-well spectrophotometer.

Western Blotting
Cells were harvested and put into RIPA lysis buffer and frozen overnight. After defrosting, the cell suspension was centrifuged at 10,000× g. The concentration of proteins was estimated by the BCA method. Laemmli SDS sample buffer was added to lysates before separation by the SDS-PAGE method. The same protein amount was loaded into each well. The semi-dry transfer was used (20 V for 20 min). After the transfer membranes were stained with Ponceau S and cut according to the protein size ladder. After that, membranes were blocked in 5% non-fat milk and probed overnight at 4 • C with antibodies specific (1:1000) for HO-1, CAT, GPx1, p21, p-S6, p-p53, p53, E-cadherin, Snail, Nanog, cyclin A, cyclin B, p-Rb, p-cdc2, and PARP-1. GAPDH (1:50,000) or Vinculin (1:1000) was used as a loading control. For the process of the incubation, antibodies were diluted in TBS buffer supplemented with: 5% Bovine Serum Albumin or 5% non-fat milk, 0.05% Tween, and 0.01% sodium azide. Then, proteins were detected using appropriate secondary HRP-conjugated antibodies (1:10,000), diluted in TBS buffer supplemented with 5% non-fat milk and 0.05% Tween, and ECL reagents as recommended by the manufacturer. The uncropped blots are shown in the Supplementary Materials.

Detection of Senescence-Associated β-Galactosidase (SA-β-Gal)
The activity of SA-β-gal was assessed according to modified Dimri et al. [96]. Cells were trypsinized and fixed with 2% formaldehyde and 0.2% glutaraldehyde diluted in PBS. After that, cells were washed with PBS, cytospined on microscopic slides, and incubated overnight at 37 • C with a solution containing 1 mg/mL 5-bromo-4-chloro-3-indolyl-b-Dgalactopyranoside, 5 mM potassium ferrocyanide, 150 mM NaCl, 2 mM MgCl 2 , and 0.1 M phosphate buffer, pH 6.0. All the reagents were brought to room temperature before they were mixed. Microscopic slides were sealed with the use of a Dako medium.

DNA Content Analysis
Cells were harvested, centrifuged at 1200× g for 5 min, suspended in 500 µL of PBS, and fixed in 5 mL of 70% frozen EtOH. Fixed cells were stored at −20 • C for at least 24 h. Before analysis, cells were centrifuged to remove EtOH. Then, cells were washed with PBS, centrifuged, and suspended in 250 µL of Muse TM Cell Cycle Reagent or Cell Cycle Analysis Kit (Sigma). DNA content analyses were performed using a Becton-Dickinson FACS Calibur and the BD CellQuest Pro 6.0 software or CytoFlex Beckman Coulter and CytExpert software. Cells were left in the dark for 30 min during the process of staining. Data for 50,000 cells were collected and analyzed.

Statistical Analysis
All experiments were repeated at least three times. The experiments to determine cell viability with the use of MTT assay were performed in three technical repetitions. The numerical results are expressed with mean values ± standard errors. p-values were assessed using the type-2 two-tailed t (Student) test. p < 0.05 was considered statistically significant.

Conclusions
In the current study, we show that oxidative stress induced by hemin or H 2 O 2 could be a trigger for escaping from the senescence of colon cancer cells. However, the outcome depends on the proper activities of antioxidative enzymes: HO-1, CAT, and GPx-1. Therefore, a proper combination of prooxidants and/or antioxidative enzymes' activity could prevent senescent cells from re-activating their proliferative properties and cancer re-growth after chemotherapies.  Data Availability Statement: The datasets generated during and analyzed during this study are available from the corresponding author upon reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.                . Cell cycle analysis was performed with the Cell Cycle Analysis Kit (Sigma) and flow cytometry; Each bar represents mean ± SEM, n ≥ 3; * p < 0.05-vs. control normoxia; # p < 0.05, ### p < 0.001-vs. control hypoxia; $ p < 0.05, $$ p < 0.01-normoxia vs. hypoxia.