Abstract
Background/Aim: Fisetin is a yellow-coloring flavonoid that can be found in a wide variety of plants, vegetables, and fruits, such as strawberries, apples, and grapes. It has been shown to have biological activity by targeting different pathways regulating survival and death and to bear antioxidant and anti-inflammatory activity. Fisetin was shown to be cytotoxic on different cancer cell lines and has the ability to kill therapy-induced senescent cancer cells. The aim of the study was to investigate the DNA damaging and cytotoxic potential of fisetin and its ability to enhance the killing effect of temozolomide on glioblastoma cells. Materials and Methods: We used LN229 glioblastoma cells and measured survival and apoptosis by flow cytometry, DNA strand breaks by the alkaline comet and γH2AX assay, and the DNA damage response by western blot analysis. Results: Fisetin was cytotoxic on glioblastoma cells, inducing apoptosis. In the dose range of 40-80 μM it also induced DNA damage, as measured by the alkaline comet and γH2AX assay, and triggered DNA damage response, as revealed by p53 activation. Furthermore, fisetin enhanced the genotoxic effect of methyl methanesulfonate, presumably due to inhibition of DNA repair processes. When administered together with temozolomide, the first-line therapeutic for glioblastoma, it enhanced cell death, reduced the yield of senescent cells following treatment and exhibited senolytic activity on glioblastoma cells. Conclusion: Data show that high-dose fisetin has a genotoxic potential and suggest that, harnessing the cytotoxic and senolytic activity of the flavonoid, it may enhance the effect of anticancer drugs and eliminate therapy-induced senescent cells. Therefore, it may be useful for adjuvant cancer therapy, including glioblastoma, which is worth to be studied in clinical trials.
Fisetin (3,3′,4′,7-tetrahydroxyflavone) is a flavonoid present at variable amounts in several fruits and vegetables, such as strawberries, apples, kaki, onions, and many others. Especially high amounts are found in the wood of the purple smoke bush (Cotinus coggygria) (1). Like many other phytochemicals and plant polyphenols, it bears antioxidative properties (2). Moreover, it has been reported to possess anti-viral, antibacterial and anti-inflammatory activities (3). In vivo studies with rats revealed that fisetin is able to enhance the expression levels of the antioxidant enzymes catalase (CAT) and superoxide dismutase (SOD), resulting in a decrease in the level of intracellular reactive oxygen species (ROS) (4), which explains at least in part the antioxidative effects. A recent study with the worm Caenorhabditis elegans has aroused great interest, demonstrating extension of mean and maximum life span, attenuation of age-dependent motility and inhibition of degeneration of dopaminergic neurons. The effects were mediated by DAF-16, which is involved in stress response and autophagy (5).
Fisetin was reported to modulate multiple cell targets and was studied in different cell types and model systems. Thus, besides DAF-16, kinases involved in the PI3K, mTOR and ERK/JNK pathways were shown to be inhibited by the flavonoid, while the NF-kB and AMPK pathways became activated (2). Fisetin also targets the Nrf2 regulated heme oxygenase 1 (HO-1) antioxidant system and attenuates hydrogen peroxide induced DNA damage and apoptosis (6), supporting its role in ROS protection.
Given these biochemical properties, it is not surprising that fisetin has been extensively studied as to its health-promoting effects in humans (7). Thus, fisetin has been demonstrated to have a potential in attenuating Alzheimer’s and Parkinson’s disease (8) and was reported to be beneficial in the treatment of ischemic stroke (9). It also shows beneficial effects on neurodegenerative diseases and is able to restore cognitive and memory impairments (10).
Most interestingly, fisetin was shown to increase the lifespan of mice upon dietary supplementation (11). This is likely a result of its senolytic activity, i.e., selective killing of senescent cells that accumulated in old tissues (12, 13). The studies on mice revealed that fisetin was a most potent inhibitor of cellular senescence when compared to other plant polyphenols (11). The senolytic activity of fisetin was confirmed by in vitro studies, where fisetin was shown to eliminate senescent cells, while leaving proliferating cells unaffected (11). Hence, fisetin is considered to be a potent natural senolytic agent. Because of the beneficial effects in neuroprotection and life extension, the flavonoid gained popularity as food supplement and nutraceutical, and is today worldwide commercially available over the counter.
Interestingly, in several experimental cancer systems fisetin was shown to bear cytotoxic activity. Thus, cell death was provoked by fisetin in different cancer cell lines, including those derived from lung, colon, prostate and pancreatic cancer as well as malignant melanoma (2). Therefore, a role of fisetin in cancer treatment is being exploited extensively [reviewed in (14)], and several studies indicate that the anti-proliferative and apoptosis-inducing effect of fisetin pertains specifically to cancer cells (15). Recently, in triple-negative breast cancer cells fisetin was shown to induce DNA double-strand breaks (DSBs), reduce colony formation and enhance the radiation effect due to inhibition of DNA repair functions (16), indicating fisetin might be beneficial in breast cancer therapy. To our best knowledge, the genotoxic activity of fisetin together with cell death has not been tested in malignant glioma cells.
The most severe form of brain cancer, glioblastoma (GBM), has an incidence rate of up to 5.0 per 100,000 people-years (17). Although this is still low, they are constantly increasing. Since the 1990s, when incidence rates were about 2.5 per 100,000 people-years, they have almost doubled (18). Although the incidence is low compared to other tumors, GBM is the most common malignant brain tumor, accounting for 54-57% of all gliomas (19). Classified as WHO grade 4 tumor, GBM belongs to the most invasive and most aggressive brain cancers. Recurrences are common and occur about 5 months following resection of the tumor. This leads to a dismal prognosis with a 5-year survival rate of less than 6% and a median survival of 15 months (20). The therapy involves maximal tumor resection followed by radiochemotherapy (21). The chemotherapeutic agent used first-line for GBM is temozolomide (TMZ), a DNA methylating genotoxic agent. TMZ induces apoptosis, autophagy, and senescence in GBM cells (22, 23). The pathways activated by TMZ and triggering these responses are well described. They include DSB formation due to futile mismatch repair on O6-methylguanine-thymine mismatches, activation of the downstream ATR/ATM-CHK1/CHK2 damage response cascade resulting in SIAH1-HIPK2 dependent p53 activation, and activation of apoptotic pathways (24). At the same time, the DNA damage triggers p16/p21-dependent senescence (25). Recently, we showed that in glioblastoma cells (not expressing the repair protein MGMT) senescence is the main trait induced by the critical primary lesion O6-methylguanine, while only a minority of cells undergo cell death by apoptosis (23). Based on the data we proposed that the high rate of senescence might be a contributing factor for the low curative response of patients following TMZ treatment. In view of the inefficiency of therapy, other treatment options aimed at increasing the cell death rate and also eliminating senescent cells are highly desirable.
Of particular interest are plant ingredients that could be used to support conventional therapy with radiation and genotoxic therapeutics such as temozolomide. One of these might be fisetin, which bears not only cytotoxic but also senolytic activity. As it is widely used as food supplement without reported side effects, it can be considered safe at least in a particular concentration range. Herein, we investigated the cytotoxicity and the genotoxic potential of fisetin using glioblastoma cells. We also studied the ability of fisetin to reduce senescence induction following TMZ. Together with our previous report on the senolytic activity of fisetin on glioblastoma cells (26), the data support the use of fisetin as a supplement in cancer therapy.
Materials and Methods
Cell culture and treatments. The human glioblastoma cell line LN229 (Research resource identifier (RRID): CVCL_0393) was obtained from the American Type Culture Collection (ATCC). The cells were cultured in a 5% CO2 humidified atmosphere at 37°C in Dulbecco’s Modified Eagle Medium (DMEM) GlutaMax (Gibco, Thermo Fisher Scientific Inc, Waltham, MA, USA), supplemented with 10% fetal calf serum (FCS, Gibco, Thermo Fisher Scientific Inc). For the experiments, the cells were seeded into either 10- or 5-cm cell culture dishes or 6-, 12-, 24-, or 96-well plates (Greiner BioOne GmbH, Frickenhausen, Germany, or TPP Techno Plastic Products AG, Trasadingen, Switzerland). They were treated at about 70% confluency to ensure proliferation throughout the experiments. Fisetin was from Abcam (Ab 142429; CAS-No. 528-48-3, Cambridge, UK), dissolved in ethanol and stored as stock solution at −20°C. It was diluted in PBS immediately before use and added to the medium of exponentially growing cells (not earlier than 2 days after seeding of 105 cells per 5-cm dish) to the desired final concentration and kept on the cells until harvest. Temozolomide treatment and generation and treatment of the senescent cell population was done as previously described (26).
Cell death, apoptosis, and senescence. MTT assay was performed as described previously (27). Apoptosis and late apoptosis/necrosis levels were quantified by flow cytometry. In brief, cells were collected via trypsinization, washed, and stained with FITC-coupled annexin at a concentration of 5 μl/ml in PBS at RT for 15 min and thereafter with propidium iodide (PI), 50 μl/ml in PBS, on ice for 5 min. The measurement was performed using the BD FACS Canto II and FACS Diva software (Becton Dickinson GmbH, Heidelberg, Germany) and cells were quantified using the flowing software 2 (Perttu Terho, Turku Biosciences, Turku, Finland). For the differentiation between living cells and early and late apoptosis/necrosis, cells were separated by A/PI double negative (=living cells), A positive/PI negative (=early apoptotic cells) and by A/PI double positive (=late apoptotic/necrotic cells). Cellular senescence was induced and determined by flow cytometry using C12FDG (Abcam, Cambridge, UK) as previously described (23). Alkaline comet assay. DNA single-strand breaks (SSBs) were measured using the alkaline comet assay. Additionally, oxidative damage was measured using the formamidopyrimidine-DNA glycosylase (FPG) modified alkaline comet assay. Therefore, cells were collected by trypsinization, washed in cold PBS, and resuspended in 0.5% low melting point agarose in PBS. They were then spread onto microscope slides covered in 1.5% agarose and let air-dry at 4°C under a coverslip. After polymerization, the coverslip was removed, and the cells were lysed in 4°C pre-chilled alkaline lysis buffer for 50 min. For the FPG modification, cells were thereafter first incubated for 5 min at RT in FPG-buffer, and then in a 1 μg/ml FPG solution for 45 min at 37°C under a coverslip. For single-cell gel electrophoresis (SCGE), the coverslips were removed, and the slides were transferred to the electrophoresis chamber, filled with alkaline electrophoresis buffer, and incubated for 20 min at 4°C. The SCGE was run at 300 mA for 15 min at 4°C. Following the SCGE, cells were briefly washed in distilled water, incubated for 5 min at RT in 100% ethanol and let air-dry. For detection of the comets, cells were stained with a 50 μg/ml PI solution under the coverslip, and comets were assessed using a fluorescence microscope and the Comet IV software.
γH2AX assay. For quantification of γH2AX foci, cells were seeded onto coverslips. Before staining cells were washed with cold PBS twice, and then dehydrated and permeabilized in a methanol and acetone buffer (7:3) for 6 min at −20 °C. For fixation, the cells were incubated in 4 % paraformaldehyde (PFA) in PBS for 10 min at RT. Before incubation with the primary antibody, the residual PFA was removed, and cells were re-hydrated by incubation with cold PBS three times for 5 min each, and then blocked in 3 % BSA in PBS overnight at 4°C. Incubation with the primary γH2AX antibody (9718S, Cell Signaling Technology Corporation, Danvers, MA, USA) occurred at a 1:1,000 dilution in 0.3% Triton X-100 in PBS overnight at 4°C. Cells were then washed with PBS three times for 5 min each and then incubated with the secondary antibody (ab150077, Abcam) at a 1:500 dilution in 0.3% Triton X-100 in PBS for 1h at RT in the dark. The residual secondary antibody was removed by three washing steps with PBS for 5 min each. To stain the nucleus, cells were incubated with TO-PRO™-3 Iodide (Invitrogen, Carlsbad, CA, USA) for 5 min. The stained cells were then mounted onto microscope slides using the Vectashield mounting solution (Vector Laboratories, Newark, CA, USA). To inhibit de-hydration, the slides were sealed with clear nail polish. The γH2AX foci were assessed using the LSM 710 (Carl Zeiss GmbH, Jena, Germany) and the Zeiss ZEN imaging software 2.1 (Carl Zeiss GmbH). Quantification of the foci was performed using the ImageJ software (U.S. National Institutes of Health, USA).
Western blot experiments. Whole-cell extracts were prepared by direct lysis in 1x SDS-PAGE sample buffer and subsequently sonified. After western blot transfer onto a nitrocellulose membrane, the membrane was blocked for 2h with 5% BSA, 0.1% Tween-TBS. Thereafter, the membrane was incubated overnight at 4°C with mouse mAb against γH2AX (9718, Cell Signalling Technology), p53Ser15 (9286, Cell Signalling Technology), β-Actin (sc-47778, Santa Cruz Biotechnology, Santa Cruz, CA, USA) and HSP90 (sc-13119, Santa Cruz Biotechnology), diluted 1:1,000-1:2,000 in 5% BSA, 0.1% Tween-TBS. After 3 times washing in 0.1% Tween-TBS for 10 min, the membrane was incubated with secondary HRP conjugated anti mouse antibodies (KCB002, Rockland, Limerick, PA, USA, diluted 1:4,000) for 2h at room temperature. After additional washing steps, the protein-antibody complexes were detected by Pierce® ECL Western Blotting Substrate (Thermo Fisher, Dreieich, Germany).
Statistics. Data were compared by the unpaired t-test with Welch’s correction.
Results
First, we assessed the cytotoxic activity of fisetin. As shown in Figure 1, fisetin reduced the viability of glioblastoma LN229 cells when treated in the dose range of 20-100 μM for 48 h.
Effect of fisetin on proliferating LN229 glioblastoma cells. Viability (OD: 570 nm) was measured 2 d after addition of fisetin to the medium. Data are the mean of at least 3 experiments. ****p<0.0001.
Viability data determined by the MTT assay are affected by both proliferation inhibition and cell death. Therefore, we determined the cell death rates by annexinV/PI-flow cytometry. As shown in Figure 2, cell death following fisetin treatment resulted mainly from apoptosis, which increased dose-dependently. A concentration of 60 μM significantly induced apoptosis and to some extent also necrosis. Thus, a concentration of ≥40 μM is clearly cytotoxic, inducing apoptosis and necrosis.
Cell death induced by fisetin in LN229. Exponentially growing cells were treated for 48 h and harvested for A/PI measurements. Data represent the mean of at least 3 independent experiments ±SEM. **p<0.01, ***p<0.001.
Next, we investigated whether fisetin exerts DNA damaging activity. The genotoxic effect of fisetin was analyzed by means of the alkaline comet assay. As shown in Figure 3, with doses ≥40 μM the DNA damage (tail intensity) was significantly enhanced above the background, indicating that fisetin is genotoxic, at least in the dose range 40-100 μM (Figure 3). Interestingly, there was a tendency of increasing the tail intensity in the assay in the presence of FPG, which recognizes 8-oxo-guanine in the DNA and generates SSBs following base removal (Figure 3). The difference was, however, not significant. This data indicates that the genotoxic effect of fisetin is not brought about to a significant extent by oxidative DNA damage recognized by FPG.
Genotoxic effect of fisetin on LN229 cells. (A) Exponentially growing LN229 cells were treated with fisetin for 24 h and harvested for the alkaline comet assay, which was performed in the presence and absence of FPG protein. Data are the mean of at least 3 independent experiments ±SEM. *p<0.05, **p<0.01. The difference between -FPG and +FPG was not significant. (B) Representative images of cells subjected to single cell gel electrophoresis.
The potency of fisetin to induce DSBs was assessed by means of the γH2AX foci assay. As shown in Figure 4, the mean number of foci/cell and the number/proportion of cells heavily damaged was significantly elevated above the background with doses ≥40 μM. The data support the notion that fisetin is genotoxic on glioblastoma cells in this dose range.
Genotoxic effect of fisetin as measured by the γH2AX foci assay. A, LN229 cells grown on cover slips were treated for 6 or 24 h with fisetin, immediately thereafter washed in PBS and subjected for γH2AX staining. Evaluation of stained nuclei was done by LSM. Representative images are shown in panel B. At least 100 cells were counted per measure point. **p<0.01, ****p<0.0001.
To further confirm the data, we measured the γH2AX level in Western blots. As shown in Figure 5, γH2AX was clearly enhanced following treatment of cells for 24 h. We also observed a marked increase of serine 15 phosphorylation of p53, which indicates p53 activation occurring at a dose level of ≥40 μM (Figure 5).
Effect of fisetin on the level of γH2AX and p53ser15. β-Actin and HSP90 served as internal control. LN229 cells were treated for 24 h with increasing concentrations of fisetin.
The genotoxic effect of fisetin could be speculated to result from inhibition of repair processes such as base excision repair (BER), which removes spontaneously generated base damages. If true, the effect of a genotoxic agent is anticipated to be enhanced in the presence of fisetin. This was tested by treating LN229 cells with methyl methanesulfonate (MMS), which induces DNA methylations that give rise to single-strand breaks during BER (28). Fisetin (20 μM) enhanced the DNA damage level significantly in cells treated with MMS (Figure 6). This finding supports the notion in a previous publication by Khozooei et al. that fisetin is able to inhibit repair processes (16).
Effect of fisetin on LN229 cells treated with MMS. Exponentially growing LN229 cells were treated with MMS for 1 h and cells were harvested for the alkaline comet assay after a recovery time of 4 h in the absence and presence of 20 μM fisetin. Data are the mean of at least 3 experiments ±S.E.M. **p<0.01, ***p<0.001.
First-line therapeutic for glioblastoma is the methylating drug TMZ. Although it targets the same sites in the DNA like MMS, it induces a significantly higher amount of O6-methylguanine, which is, in MGMT lacking tumors, the preponderant killing lesion (29). To test whether fisetin has an impact on TMZ-induced killing, we treated glioblastoma cells with TMZ and added 1 h later fisetin. Thus, TMZ-treated cells were incubated in the absence and presence of fisetin until harvest (120 h later). As shown in Figure 7A, fisetin treated along with TMZ enhanced the yield of apoptosis and at the same time reduced the level of TMZ-induced senescence.
Effect of fisetin on the induction of senescence and apoptosis (A) and senescence maintenance (B). (A) exponentially growing LN229 cells were treated with temozolomide (50 μM) together with fisetin at a concentration of 20, 40 or 60 μM. The yield of senescent cells was measured by c12FDG and apoptosis/necrosis by A/PI flow cytometry. (B) LN229 cells in the senescence state (8 d after TMZ treatment, 50 μM) were not treated or treated with fisetin at a concentration of 40 or 60 μM. 2 days later, cells were harvested and measured by flow cytometry for c12FDG (senescence) and A/PI (apoptosis/necrosis). Data are the mean of at least 3 independent experiments ±SEM. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.
Fisetin was reported to have senolytic activity. If administered to glioblastoma cells that were treated with TMZ and arrested in the senescent state, the number of senescent cells in the population declined and the apoptotic fraction increased (Figure 7B). All the effects described above were observed in a narrow concentration window of 40-80 μM fisetin. Above this concentration, effects could not be measured anymore because of excessive cell death.
Discussion
The flavonoid fisetin, present in fruits and vegetables, is a potent senolytic agent (12). Since senescent cells accumulate in the body during normal aging, clearing of senescent cells is believed to slow down the aging process. Fisetin is, therefore, advertised as an effective antiaging supplement.
Cellular senescence is not only a hallmark of the normal ageing process, but also a by-product of cancer therapy, provoked by genotoxic anticancer drugs and ionizing radiation. Although related to severe side effects, such as fibrosis, cardiac dysfunction and macular degeneration (30), cellular senescence has been thought to be beneficial in cancer therapy as senescent cells do not proliferate and therefore tumor growth is inhibited. However, evidence is mounting that senescent cells can escape from cell cycle arrest and re-start proliferation, leading to delayed tumor progression and the appearance of recurrent tumors months or even years after treatment (31, 32). Importantly, senescent cells secrete pro-inflammatory cytokines and interleukins, which is a hallmark of the senescence-associated secretory phenotype (SASP) (33). Secretion of SASP factors into healthy and cancerous tissue leads to inflammation, re-enforcement of senescence and presumably also tumor progression (34). SASPs include cytotoxic cytokines such as TNF-alpha that can trigger cell death in normal cells, while senescent cells are equipped with an arsenal of antiapoptotic factors that make them resistant to the toxic inflammatory environment (30). In view of these negative effects of senescent cells, their elimination is a desirable goal in cancer therapy.
In glioblastoma therapy, treatment with TMZ concomitant with radiation (according to the so-called “Stupp protocol”) followed by daily adjuvant TMZ belongs to the standard repertoire after primary tumor resection (35). Thus, glioblastoma patients are exposed to TMZ daily over long periods of time, with an accumulation of the critical damage O6-methylguanine in the MGMT-lacking tumor tissue, which leads to cumulative effects regarding the endpoints apoptosis and senescence (36). An important and unexpected finding was that TMZ induces cellular senescence along with cell death through apoptosis (22). The upstream pathways of apoptosis and senescence are nearly identic. They are triggered by the critical damage O6-methylguanine and conversion of the lesion by mismatch repair into replication blocks and DSBs that activate complex downstream responses (24, 25). TMZ-induced senescent glioblastoma cells are arrested in G2, down-regulate various repair pathways, activate antiapoptotic factors and secrete proinflammatory cytokines, thus exhibiting the SASP (25). They display a high level of ROS, oxidized DNA damage, DSBs and sustained activation of the DNA damage response, i.e., activation of the ATR/ATM-CHK1/CHK2-p53-p21 axis (24). Actually, cellular senescence (CSEN) is the major trait triggered by O6-methylguanine in p53 proficient glioblastoma cells (23). Cells characterized by the CSEN phenotype were also found in recurrent tumors in situ at higher level than in the primary tumor (23), supporting the notion that CSEN is induced in vivo in the course of TMZ-based therapy. Prompted by the finding that TMZ (and radiation) efficiently induces senescence, we conducted a study aimed at identifying senolytic agents that target GBM cells. As reported previously, a screening of natural and synthetic compounds on TMZ-induced senescent GBM cells revealed fisetin as a natural compound with senolytic activity on glioma cells (26). This confirmed and extended the initial report on fisetin (12) and makes the flavonoid attractive as supplement in post-primary and adjuvant GBM therapy.
Fisetin bears not only senolytic activity. Here, we extended our previous study and show that, in a high concentration range, fisetin is cytotoxic on proliferating, non-senescent cells, causing the induction of apoptosis. This confirms an early study on glioma cells (37) and other reports, demonstrating that fisetin induces apoptosis in different cancer cell lines derived from skin and ovarian cancer (38) [for review see (14)]. It was also effective in inhibiting cell migration and invasion of GBM8401 glioma cells via ERK1/2 pathway activation. These effects occurred in the concentration range of 10 to 40 μM (39), which is close to the cytotoxic and genotoxic range in our study.
We further showed that fisetin induces DNA single-strand and DSBs. The latter confirms a recent report with triple-negative breast cancer cells, in which fisetin was active as radiosensitizer through suppression of the non-homologous end-joining (c-NHEJ) and homologous recombination (HR) repair pathways (16). We also demonstrate that fisetin enhances the genotoxic effect of MMS and ameliorates the apoptosis-inducing effect of TMZ. A key defense mechanism for alkylating agents, including MMS and TMZ, is HR, which is required for resolution of blocked replication forks and repair of DSBs resulting from them (40, 41). The increased levels of MMS-induced DNA breaks in the presence of fisetin as well as the increased apoptosis levels following TMZ plus fisetin can be explained on this basis. In line with this is the finding that inhibition of HR enhances the cytotoxic activity of TMZ in glioblastoma cells (42). Overall, the data support the notion that fisetin might be useful as supplement in concomitant TMZ/radiotherapy and adjuvant TMZ. Clinical studies are warranted.
Similar to other dietary flavonoids, fisetin appears to have both antioxidant and prooxidant properties (43). The documented antioxidant properties at low dose levels are considered to contribute to cancer protection and prevention, while the high dose prooxidant effect and genotoxicity as well as inhibition of repair processes (16) might play a decisive role in cancer treatment. The prooxidant effects are associated with the oxidative properties of flavonoids, resulting from intracellular ROS production, such as superoxide anions, hydrogen peroxide and reactive quinones (44), which damage the tumor cell’s DNA, resulting in the effects observed. It is interesting that copper can strongly promote flavonoid oxidation and formation of hydroxyl radicals (45). A copper-mediated redox reaction of flavonoids and the generation of hydroxyl radicals in the proximity of the DNA can cause DNA strand breaks that cannot be rescued by radical scavengers (46). It is important to note that many cancer cells have increased cellular copper (47) and thus it is reasonable to conclude that a selective anticancer mechanism of plant flavonoids, including fisetin, rests on the high intracellular copper concentration of cancer cells, which promotes the cytotoxicity of plant flavonoids (43). The serum copper concentration was found to be significantly higher in glioma patients compared to other neurological diseases and healthy individuals (48), which supports the notion that fisetin may target glioblastoma. Clearly, further studies, including organoids and in vivo analyses, are warranted to clarify whether fotemustine exerts a tumor-specific cytotoxic response.
Like many other plant polyphenols, fisetin has a low water solubility and, therefore, a low bioavailability. Unfortunately, data on human plasma concentrations following oral administration of fisetin are lacking. In mice, a peak concentration of 2.5 and 7 μg/ml was reached 15 min following intraperitoneal administration of 223 mg/kg BW free and 21 mg/kg BW liposomal fisetin, respectively (49). The peak concentration after liposomal fisetin corresponds to 25 μM serum concentration, which is near to the effective dose in our and previously published experiments (26). New delivery methods, such as encapsulation in lipids, are being developed for enhancing the bioavailability of the flavonoid (50). Nevertheless, mice that received liposomal fisetin intraperitoneally (21 mg/kg BW) showed a nearly constant plasma level of about 2 μg/ml over a period of 4 h. This treatment inhibited the tumor growth in the Lewis lung carcinoma model and ameliorated, in this model system, the anticancer effect of cyclophosphamide (51). Thus, liposomal administration of fisetin clearly improved the bioavailability and anticancer activity of the agent.
Fisetin embedded in liposomes was also tested together with cisplatin as an alternative/supportive strategy for glioblastoma therapy: liposomal fisetin retained the cytotoxic activity on U-87MG glioblastoma cells, as measured by the MTT assay (52). It should be noted that low water solubility is also the limiting factor of other bioactive plant polyphenols, such as curcumin. Different delivery methods were tried out here, with very good results as to bioavailability being obtained with curcumin embedded in micelles (53). Intravenous infusion of high dose curcumin solubilized in organic solvent or liposomal formulation is applied in medical practice (54). Despite encouraging results, further studies are required to optimize the bioavailability of fisetin.
Finally, we would like to discuss the aspect of genotoxicity. Since fisetin is used as a food supplement, it is important to know whether fisetin is safe under consumption conditions. Although our investigation along with reports from other groups (16, 55-57) clearly indicates that fisetin has a genotoxic potential, it is unlikely that the harmful concentrations can be reached in vivo. In animal experiments, the peak plasma level 15 min after i.p. administration of free fisetin (223 mg/kg BW) was 3 mg/mL and 4 h later 0.3 mg/ml (51), which corresponds to 10 and 1 μM, respectively. This appears to be below the genotoxic level. It should be noted that the administered concentration in the mice experiment referred to above corresponds to about 16 grams fisetin consumed by a 70 kg individual, which is far from being realistic. In food supplements fisetin is usually offered at a concentration of 500 mg/capsule, and the recommended dose is maximally 2 capsules per day. Nevertheless, measurement of serum levels following fisetin in human volunteers are desired.
In a way, fisetin resembles curcumin, which also shows genotoxic effects in vitro in high concentrations (≥40 μM) that cannot be achieved in vivo under the recommended acceptable daily intake conditions (58). It is important to note that it is unclear whether the apoptosis-inducing and genotoxic effects are limited to tumor cells or also affect normal cells of the body, including brain and other tissues such as the hematopoietic system. The present study hopefully stimulates further investigations in this direction.
Conclusions
In summary, we herein report that fisetin is cytotoxic and genotoxic in LN229 cells (concentration range 40-80 μM), which confirms and extends previous studies on other cancer cell systems. This high concentration can not be achieved if fisetin is used as food supplement. Since fisetin has low bioavailability and, at low concentration, bears antioxidative activity, the beneficial effect appears to be dominating if fisetin is used as supplement. However, in adjuvant cancer therapy, clearly higher doses are required in order to provoke prooxidative effects that are related to genotoxicity and death of cancer cells. Since apoptosis and genotoxicity are induced in the same dose range, it is reasonable to suppose that genetically damaged cells are eliminated by cell death and, therefore, long-term side effects resulting from genotoxicity on non-transformed cells may be negligible under high dose conditions.
Fisetin reduced the senescence-inducing effect of TMZ and therefore may be used supportive in adjuvant GBM therapy. Like other flavonoids, it may also bear neuroimmuno-modulatory properties, which was recently reviewed (59). By and large, fisetin is active on cancer cells in three ways: by direct elimination of cancer cells, by attenuating senescence induction, and by the elimination of drug-induced senescent cells. The low bioavailability of the flavonoid might be overcome by new delivery methods, including encapsulation in liposomes or micelles. In GBM therapy, intracerebral high dose delivery, as reported previously (60), might also be an option.
Regarding the senolytic activity on glioblastoma, we should note that similar to other senolytic agents, treatment should occur in a hit-and-run fashion, i.e., TMZ treatment cycles should be interrupted by short high-dose fisetin cycles. Fisetin is commercially available as supplement, for example with a concentration of 500 mg/capsule and a recommended daily dose of 1,000 mg. Two capsules per day taken over several months have been very well tolerated and no side effects have been observed in a glioblastoma patient post radio-chemotherapy and adjuvant TMZ and lomustine (case report, unpublished). Overall, fisetin bears a significant therapeutic potential that should not be neglected and further explored in clinical trials.
Acknowledgements
We are grateful to Anna Frumkina for her assistance in the comet assays.
Footnotes
Authors’ Contributions
BK, LB, and MC conducted the experiments, analyzed and presented the data. BK was responsible for conceptualization, data curation, supervision, funding acquisition, and writing of the original draft. LB, MC, MD, and BK analyzed and validated the data. All Authors red and approved the submitted version of the manuscript.
Conflicts of Interest
The Authors declare no conflicts of interest. The funder 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.
Funding
Work was supported by Deutsche Forschungsgemeinschaft, grant DFG KA724/31-1 to BK.
- Received October 4, 2023.
- Revision received February 1, 2024.
- Accepted February 2, 2024.
- Copyright © 2024 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
