Xanthohumol, a Prenylated Flavonoid from Hops, Induces DNA Damages in Colorectal Cancer Cells and Sensitizes SW480 Cells to the SN38 Chemotherapeutic Agent.

In spite of chemotherapy and systematic screening for people at risk, the mortality rate of colorectal cancer (CRC) remains consistently high, with 600,000 deaths per year. This low success rate in the treatment of CRC results from many failures associated with high resistance and the risk of metastasis. Therefore, in response to these therapeutic failures, new strategies have been under development for several years aimed at increasing the effect of anticancer compounds and/or at reducing their secondary effects on normal cells, thus enabling the host to better withstand chemotherapy. This study highlights that xanthohumol (Xn) concentrations under the IC50 values were able to induce apoptosis and to enhance the DNA-damage response (DDR). We demonstrate for the first time that Xn exerts its anticancer activity in models of colon cancer through activation of the ataxia telangiectasia mutated (ATM) pathway. Subsequently, the ability of Xn to restore DNA damage in CRC cells can sensitize them to anticancer agents such as SN38 (7-ethyl-10-hydroxycamptothecin) used in chemotherapy.


Introduction
In spite of chemotherapy and systematic screening for people at risk, the mortality rate of colorectal cancer (CRC) remains consistently high, with 600,000 deaths per year. The 5-year survival rate ranges from 90% in patients with stage I to 10% in patients with stage IV disease [1]. This low success rate in the treatment of CRC results from many failures associated with high resistance and the risk of metastasis. Indeed, despite the chemotherapeutic treatments used as monotherapy or in combination, as surgery/radiotherapy, and as targeted therapies, the low survival rate is partly explained by the emergence of intrinsic or acquired resistance mechanisms in tumor cells against therapeutic agents. Several major factors are usually described as being involved in the emergence of resistance to chemotherapy that may be linked to the tumor microenvironment, the host, or the cancer cell. Indeed, we have recently been able to show that tumor metabolism could play a preponderant role in the progression of CRC as well as in the emergence of resistance phenomena linked to the accumulation of complex lipids such as phospholipids [2,3]. This particular tumor microenvironment, associated with a dysfunction in the exacerbated tumor lipid metabolism, is then linked to a decrease Considering the potential of Xn as a chemopreventive agent, we investigated its ability to inhibit the proliferation of three colorectal cell lines and to induce their death by apoptosis. To determine whether Xn exerts a potential adjuvant action with chemotherapeutic drugs, we first analyzed its ability to inhibit the proliferation of three CRC cell lines. Some reports have previously described a potential antiproliferative property for Xn that is highly dependent on cell lines, times of treatment, and concentrations of this prenylated chalcone [19,28,29]. For example, Xn at 10 µM inhibited cell proliferation in the thyroid cancer TPC-1 cell line, supporting a potential action against carcinogenesis, while approximately 100 µM Xn decreased cell viability and the main proapoptotic process [30]. We highlighted that Xn concentrations under the IC 50 values were able to induce apoptosis and to enhance the DDR. We demonstrated for the first time that Xn exerts its anticancer activity in models of colon cancer by activating the ataxia telangiectasia mutated (ATM) pathway. Subsequently, the ability of Xn to restore DNA damage in CRC cells can sensitize these to anticancer agents such as SN38 (7-ethyl-10-hydroxycamptothecin) used in chemotherapy.

Cell Lines
Human colorectal cancer cell lines SW620, SW480, and HT29 were purchased from the American Type Culture Collection (ATCC, Molsheim, France). SW480 cells are derived from a Duke's B primary colon adenocarcinoma, and their metastasis-derived counterpart, SW620 cells, are derived from a colorectal adenocarcinoma Duke's C lymph node metastasis. HT29 cells are derived from a Duke's C primary colon adenocarcinoma. All cell lines have a microsatellite stable (MSS) phenotype. SW480 and SW620 cells harbor KRAS-g12V (substitution-missense) and TP53 (pR273H; P309S) mutations but are expressing wild-type (wt) BRAF, PIK3CA, and PTEN genes. HT29 cells are wt for KRAS, PI3KCA and PTEN genes but are BRAF (V600E) and TP53 (pR273H) mutated [31]. Cells were maintained in a 5% CO 2 humidified atmosphere at 37 • C and cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fetal bovine serum (Dutscher, Brumath, France). All cell lines were routinely tested for mycoplasma contamination using the Mycoalert Mycoplasma Detection Kit (Lonza, Levallois-Perret, France).

Cell Viability Assays
SW620, SW480, and HT29 cells were seeded into 12-well plates and incubated for 24 h. Cells were then treated with increasing Xn concentrations for 24, 48, and 72 h. Cell viability was first determined using trypan blue staining. The number of viable cells was counted using a KOVA Glasstic Slide (Thermo Fisher Scientific, Illkirch-Graffenstaden, France). Cell viability was also assessed by crystal violet staining (Sigma Aldrich, St. Quentin Fallavier, France). Briefly, cells were seeded into 96-well plates and treated after 24 h of culture. At the end of treatment, cells were washed with phosphate-buffered saline (PBS), fixed with ethanol for 10 min at 4 • C, and then stained with a crystal violet solution (0.5% (w/v) crystal violet in 25% (v/v) methanol) for 15 min at room temperature. Cells Cells 2020, 9, 932 4 of 20 were then gently rinsed with water, and absorbance was measured at 590 nm using a Biochrom Assays UVM 340 microplate reader, following extraction of the dye using an acetic acid 33% solution. The inhibitory concentration 50% (IC 50 ), defined as the concentration of Xn that inhibits cell proliferation by 50%, were calculated using a four-parameter nonlinear regression with GraphPad Prism version 6 software (GraphPad Software, La Jolla, San Diego, CA, USA).

Combination Index Analysis
SW480 cells were seeded into 96-well plates. After overnight incubation, cells were pretreated for 24 h with increasing Xn concentrations (starting concentration, 100 µM; four two-fold serial dilutions). The medium was then removed and replaced with fresh medium containing increasing SN38 concentrations (starting concentration, 25 µM; eight two-fold serial dilutions) for 48 h before crystal violet staining as described above. Drug interactions were quantified using the Chou-Talalay method [32,33]. Thanks to data of drug combination and single-drug inhibition experiments (dose-effect curves), the synergism, additivity, and antagonism of the different combinations of the two drugs were calculated using CompuSyn version 1.0 software (ComboSyn, Inc. Paramus, NJ, USA), on the basis of the multiple drug effect equation and quantification by the combination index (CI). CI was quantified with the following equation: CI = (D) 1 /(D x ) 1 + (D) 2 /(D x ) 2 , where D 1 and D 2 represent the doses of drugs 1 and 2 in combination required to achieve the same x% of inhibition as that of drugs 1 (D x ) 1 and 2 (D x ) 2 when used alone. CI values were generated over a range of Fraction Affected level (FA), indicating x% of inhibited cells, from 0.05 to 0.90. CI > 1, CI = 1, and CI < 1 indicate antagonistic, additive, and synergistic effects, respectively. CI values for each fixed-dose combination of the two drugs were plotted against effect levels (FA). Drug synergy was also assessed by a normalized isobologram, drawn by plotting on both the x-and y-axes the normalization of the dose of each drug (D) in combination (inducing x% inhibition) with the dose at x% inhibition (D x ) of drugs alone. Combination data points that fell on the line represent an additive interaction, whereas data points below or above the line represent synergism or antagonism, respectively. Besides synergistic combination studies, dose-reduction studies of each drug can be simulated as the inverted terms of the CI equation correspond to the dose-reduction index (DRI), where (DRI) 1 = (D x ) 1 /(D) 1 and (DRI) 2 = (D x ) 2 /(D) 2 . (DRI) 1 (or (DRI) 2 ) is defined as the fold of drug 1 (or 2) dose reduction allowed by the synergistic combination while maintaining the same % inhibition as drug 1 (or 2) alone (thereby reducing toxicity). DRI > 1 and DRI < 1 indicate, respectively, favorable and unfavorable dose reduction; DRI = 1 indicates no dose reduction. Similar to the FA-CI plot, the FA-DRI plot is generated using CompuSyn (ComboSyn, Inc., Paramus, NJ 07652, USA).

Flow Cytometry
All flow cytometry experiments were conducted on a BD FACSCanto TM cytometer equipped with BD FACSDiva software (BD Biosciences, Le Pont de Claix, France), and data were analyzed using FlowJo version 10 software (Tree Star, Ashland, OR, USA).

Cell Cycle Analysis
Cell distribution in cell cycle phases was assessed using the DNA-binding dye propidium iodide (PI). SW620, SW480, and HT29 cells were plated in six-well plates and allowed to recover overnight. Cells were then treated with 2.5, 5, and 10 µM of Xn for 24 and 48 h. Briefly, floating cells and adherent cells were collected, washed with cold PBS, and then fixed/permeabilized with ice-cold 70% ethanol for 24 h at −20 • C. After rehydration in PBS, cells were washed with cold PBS and then incubated with RNase A (200 µg/mL) and PI (50 µg/mL) (Sigma-Aldrich, St. Quentin Fallavier, France) for 1 h at 37 • C in the dark. After staining, cells were washed in cold PBS and the cellular DNA content in the different phases of the cell cycle was then analyzed with the flow cytometer.

Apoptosis Analysis
Cell viability was determined using annexin V-FITC/7-aminoactinomycin D (7AAD) staining from BD Biosciences according to the manufacturer's instructions. SW620, SW480, and HT29 cells were seeded in 12-well plates and allowed to attach overnight. The following day, cells were treated with 2.5, 5, and 10 µM of Xn for 24 and 48 h. Floating cells and adherent cells were harvested, washed twice with cold PBS, and then incubated with 5 µL of annexin V-FITC and 5 µL of 7AAD in 50 µL of binding buffer for 15 min at room temperature in the dark. At the end of the incubation, 200 µL of binding buffer was added and cells were kept on ice until analysis. Compensation controls included unstained cells and cells stained with either annexin V-FITC or 7AAD alone.

Statistical Analysis
Data are represented as mean ± standard deviation (SD) or standard error of the mean (SEM) of at least three independent experiments. Statistical analysis was conducted using GraphPad Prism version 6 software (GraphPad Software, La Jolla, San Diego, CA, USA). The comparison of continuous data was performed using one-way or two-way ANOVA as appropriate followed by Tukey's multiple comparison test after having checked data for normal distribution and variance homogeneity. All p values are two-tailed; p values < 0.05 were considered significant (* p < 0.05, ** p < 0.01, and *** p < 0.001).

Antiproliferative Properties of Xanthohumol on Colorectal Cancer Cells
We firstly evaluated the effects of Xn in thee human (SW480, SW620, and HT29) cellular models of CRC with different histological grades. Cells were firstly exposed to increasing concentrations of Xn (0.1-85 µM) for 24, 48, and 72 h prior to cell viability assays ( Figure 1B). As shown by crystal violet, which is a nonenzymatic-based assay, since it is only dependent on DNA of live adherent cells, Xn displayed a decrease in cell viability in all cell lines tested in a time-and concentration-dependent manner. It appears that the three CRC cell lines are relatively resistant to Xn after 24 h of treatment up to high concentrations, since we were able to determine the concentration inhibiting 50% of cell viability (IC 50 ) for only HT29 cell lines for tested concentrations ( Table 1). The percentage of cell viability strongly decreases after 48 h of treatment and, based on IC 50 determination at 48 and 72 h of treatment (Table 1), SW620 and HT29 cells appeared to be more sensitive to Xn than SW480 cells. Indeed, SW480 cells present an IC 50 of 10 µM higher than that of the two other cell lines after 48 and 72 h of treatment. Although crystal violet staining remains a reliable, accurate, and rapid method for the indirect estimation of viable cells and the screening of the anticancer activity of compounds, we complemented these experiments by measuring colon cancer cell growth using cell quantification by the trypan blue exclusion test, which is based on the ability of viable cells with an intact membrane to exclude the dye trypan blue by using a hemocytometer in microscopic counting ( Figure 1C). For this experiment, we have chosen four concentrations that are below the IC 50 after 72 h of Xn treatment (2.5, 5, and 10 µM) and one which is above the IC 50 (30 µM). As previously, 24 h of treatment with Xn are not sufficient to generate significant differences in the number of viable cells; however, we observed that 48 h of Xn treatment induces a significant inhibition of cell proliferation ( Figure 1C). Interestingly, Xn induced strong inhibition of cancer cell growth in a time-dependent manner from 5 µM in the three cell lines tested ( Figure 1C). However, at a concentration higher than the previously determined IC 50 , such as 30 µM, Xn appeared to be toxic, causing a decrease in the cell number. Therefore, we used nontoxic concentrations for the prenylated chalcone, Xn, for the following experiments, namely 2.5, 5, and 10 µM of Xn.   Viable cells were distinguished by trypan blue exclusion. The data are mean ± SD of three independent experiments with n = 6. p values were determined by one-way ANOVA followed by Tukey's multiple comparison test. * p < 0.05, ** p < 0.01, and *** p < 0.001.

Xanthohumol Modulates Distribution of Colorectal Cancer Cells in the Cell Cycle
Usually, tumoral growth cell inhibition is associated with disruption of normal cell cycle progression. Indeed, various studies report that checkpoints at both G1/S and G2/M of the cell cycle are found to be perturbed by phytochemicals [34][35][36] related to their antiproliferative activities. In order to determine whether Xn has the same effect, we analyzed the distribution of CRC cells in the cell cycle by flow cytometry. Colon cancer cells were exposed for 24 and 48 h to three concentrations of Xn (2.5, 5, and 10 µM), for which inhibition of cell proliferation was due to the specific effect of Xn and not due to their toxic effect ( Figure 1B,C). As observed previously for cell viability, only SW620 and HT29 present a slight modulation of the distribution of tumor cells within the different phases of the cell cycle after 24 and 48 h of treatment with 5 and 10 µM Xn ( Figure 2). Indeed, quantitative analyses of cell distribution in the cell cycle phase show a slight increase in SW620 cells in S phase from 24 h with 5 and 10 µM, which intensifies after 48 h of treatment ( Figure 2). This increase in cell number in the DNA replication phase is associated, as investigated, with a decrease in SW620 cells in G1 phase. HT29 cells also show a disturbance in their distribution in the different phases of the cell cycle, but surprisingly, instead of accumulating in the S phase of the cell cycle as for SW620, Xn induces a decrease in HT29 cells in the S phase of the cell cycle after 24 h of treatment, this time being compensated by an increase in cells in phase G1 of the cell cycle ( Figure 2). For the cells most resistant to the antiproliferative action of Xn, SW480, it can be seen that there is no change, however slight, in the distribution of cells in the different phases of the cell cycle whatever the treatment time or the concentration used ( Figure 2). Cells 2020, 9, x FOR PEER REVIEW 8 of 23 Viable cells were distinguished by trypan blue exclusion. The data are mean ± SD of three independent experiments with n = 6. p values were determined by one-way ANOVA followed by Tukey's multiple comparison test. *p < 0.05, **p < 0.01, and ***p < 0.001.

Xanthohumol Modulates Distribution of Colorectal Cancer Cells in the Cell Cycle
Usually, tumoral growth cell inhibition is associated with disruption of normal cell cycle progression. Indeed, various studies report that checkpoints at both G1/S and G2/M of the cell cycle are found to be perturbed by phytochemicals [34][35][36] related to their antiproliferative activities. In order to determine whether Xn has the same effect, we analyzed the distribution of CRC cells in the cell cycle by flow cytometry. Colon cancer cells were exposed for 24 and 48 h to three concentrations of Xn (2.5, 5, and 10 µM), for which inhibition of cell proliferation was due to the specific effect of Xn and not due to their toxic effect ( Figure 1B,C). As observed previously for cell viability, only SW620 and HT29 present a slight modulation of the distribution of tumor cells within the different phases of the cell cycle after 24 and 48 h of treatment with 5 and 10 µM Xn ( Figure 2). Indeed, quantitative analyses of cell distribution in the cell cycle phase show a slight increase in SW620 cells in S phase from 24 h with 5 and 10 µM, which intensifies after 48 h of treatment ( Figure 2). This increase in cell number in the DNA replication phase is associated, as investigated, with a decrease in SW620 cells in G1 phase. HT29 cells also show a disturbance in their distribution in the different phases of the cell cycle, but surprisingly, instead of accumulating in the S phase of the cell cycle as for SW620, Xn induces a decrease in HT29 cells in the S phase of the cell cycle after 24 h of treatment, this time being compensated by an increase in cells in phase G1 of the cell cycle ( Figure 2). For the cells most resistant to the antiproliferative action of Xn, SW480, it can be seen that there is no change, however slight, in the distribution of cells in the different phases of the cell cycle whatever the treatment time or the concentration used (Figure 2). Histograms represent quantitative analysis of cell distribution in the cell cycle phases (mean ± SD of three independent experiments). p values were determined by a two-way ANOVA followed by Tukey's multiple comparison test. * p < 0.05, ** p < 0.01, and *** p < 0.001.

Involvement of Cyclins and their Cdks in Xanthohumol-Induced Cell Cycle Disruption
Modulation of tumor cell distribution in the different cell cycle phases prompted us to investigate whether the key regulators of cell cycle progression, cyclins and their partners, the cyclin-dependent kinases (Cdks) could be affected upon exposure to Xn. We have previously shown with another natural compound-resveratrol, an emblematic polyphenol-that polyphenol-induced colon cancer Cells 2020, 9, 932 9 of 20 cell accumulation in the S phase was associated with an increase in cyclin A and B expression as well as Cdk2 protein levels, which have been shown to be required for DNA synthesis [12,16,37,38]. We observed similar results with Xn, where a slight increase is observed at 24 h of treatment with Xn concerning Cdk2 and cyclin E in SW620 cell line ( Figure 3). Then, at 48 h of treatment, Xn increases cyclin A and E expression, whereas cyclin B is slightly decreased at the higher concentration of 10 µM (Figure 4). Consistent with the entry of cells into the S phase, there is usually an increase in nuclear Cdk2 activity associated with both cyclins A and E [16,39]. Conversely, the G0/G1 phase is triggered by the cyclin E that appears at the start of the G1 phase and reaches its maximum level near the G1/S border. Cyclin E associates with Cdk2 or Cdk1 and controls the checkpoint G1/S transition. HT29 cells exposed to increased concentrations of Xn show an increase in cyclin E as well as its associated Cdks, Cdk1 and 2, from 24-48 h of treatment (Figures 3 and 4). In accordance with results of cell viability and cell cycle distribution, the most resistant colon cancer cells, SW480, did not show modulation of the key cell cycle regulators, whether cyclins or Cdks (Figures 3 and 4).   Figure 3A. Data are expressed as mean fold induction ± SEM of three independent experiments. p values were determined by a one-way ANOVA followed by Tukey's multiple comparison test. * p < 0.05, ** p < 0.01, and *** p < 0.001.  Figure 3A. Data are expressed as mean fold induction ± SEM of three independent experiments. p values were determined by a one-way ANOVA followed by Tukey's multiple comparison test. * p < 0.05, ** p < 0.01, and *** p < 0.001.  Figure 4A. Data are expressed as mean fold induction ± SEM of three independent experiments. p values were determined by a one-way ANOVA followed by Tukey's multiple comparison test. * p < 0.05, ** p < 0.01, and *** p < 0.001.

Xanthohumol Induces DNA Damage in Colon Cancer Cells
Various reports have correlated the cell cycle delay and cell growth arrest to the induction of cell death through the induction of DNA damage and an increase in p53 phosphorylation. Both processes are reported to depend on the DNA damage checkpoint kinases ATM/ataxia telangiectasia-Rad3related (ATR) [40] and checkpoint kinase (Chk) [41]. ATM is a protein that is well known as a central mediator of responses to DNA DSB and subsequent replication stress, whereas the ATR protein is activated in response to stress at the level of replication forks and to DNA single-strand break (SSB) lesions. In all cases, ATM/ATR activation phosphorylates downstream the histone H2AX (γ-H2AX), which constitutes a recognized marker of DNA damage.
In all colon cancer cell lines tested, Xn at very low concentrations of 2.5-10 µM induces a strong increase in phosphorylated-histone H2AX (γ-H2AX), which reflects DNA damage levels at 24 h ( Figure 5) and 48 h ( Figure 6). Very interestingly, SW480, which is a more resistant cell line, exhibits strong induction of H2AX serine-19 phosphorylation from 2.5 to 10 µM Xn, which could be associated with apoptosis of the lethally damaged cells. Subsequently, the DDR pathway induced by the activation of ATM/ATR proteins is modulated in colon cancer cells. Indeed, Xn induces ATM  Figure 4A. Data are expressed as mean fold induction ± SEM of three independent experiments. p values were determined by a one-way ANOVA followed by Tukey's multiple comparison test. * p < 0.05, ** p < 0.01, and *** p < 0.001.

Xanthohumol Induces DNA Damage in Colon Cancer Cells
Various reports have correlated the cell cycle delay and cell growth arrest to the induction of cell death through the induction of DNA damage and an increase in p53 phosphorylation. Both processes are reported to depend on the DNA damage checkpoint kinases ATM/ataxia telangiectasia-Rad3-related (ATR) [40] and checkpoint kinase (Chk) [41]. ATM is a protein that is well known as a central mediator of responses to DNA DSB and subsequent replication stress, whereas the ATR protein is activated in response to stress at the level of replication forks and to DNA single-strand break (SSB) lesions. In all cases, ATM/ATR activation phosphorylates downstream the histone H 2 AX (γ-H 2 AX), which constitutes a recognized marker of DNA damage.
In all colon cancer cell lines tested, Xn at very low concentrations of 2.5-10 µM induces a strong increase in phosphorylated-histone H 2 AX (γ-H 2 AX), which reflects DNA damage levels at 24 h ( Figure 5) and 48 h (Figure 6). Very interestingly, SW480, which is a more resistant cell line, exhibits strong induction of H 2 AX serine-19 phosphorylation from 2.5 to 10 µM Xn, which could be associated with apoptosis of the lethally damaged cells. Subsequently, the DDR pathway induced by the activation of ATM/ATR proteins is modulated in colon cancer cells. Indeed, Xn induces ATM activation through its phosphorylation, which is surprisingly more important in SW480 cells compared to SW620 and HT29 cells (Figures 5 and 6). This enhancement of the phosphorylated form of ATM is associated with a decrease in ATM protein at 48 h ( Figure 6). Compared to SW620 and HT29 cells, only in SW480 cells was Xn able to induce phosphorylation of both ATM and ATR at 24 h of treatment ( Figures 5  and 6), suggesting that this prenylated chalcone could lead to DSB as well as SSB in this tumoral cell line. In all cases, whatever the form initially activated, the consequence of these activations is p53 activation, which has been shown to respond to DNA damage by an increase in its phosphorylation at serine-20 ( Figures 5 and 6). In agreement with the activation of p53, we observed that Xn induced p21 expression, which has been described as one of the transcriptional targets of the tumor suppressor protein p53 (Figures 5 and 6).     Figure 6A. Data are expressed as mean fold induction ± SEM of three independent experiments. p values were determined by a one-way ANOVA followed by Tukey's multiple comparison test. * p < 0.05, ** p < 0.01, and *** p < 0.001.  Figure 6A. Data are expressed as mean fold induction ± SEM of three independent experiments. p values were determined by a one-way ANOVA followed by Tukey's multiple comparison test. * p < 0.05, ** p < 0.01, and *** p < 0.001.

Xanthohumol Induces Apoptosis in Colon Cancer Cells
Following DNA damage, cells either repair the damage or undergo apoptotic death, mostly in a p53-dependent manner. To clarify how colon cancer cells respond to DNA damage, we investigated Xn-induced cell death. Colon cancer cells were double stained with annexin V/7AAD after treatment with Xn at 2.5, 5, and 10 µM for 24 h and 48 h. It appears that Xn induces an increase of global level of apoptosis in the three tested colorectal cancer cell lines as compared to the control (Figure 7). More particularly, the level of early apoptosis is more important than late apoptosis in HT29 cells, which starts from 24 h of treatment in a concentration-dependent manner compared to the control ( Figure 7A,B).

Xanthohumol Synergized with SN38 to Inhibit SW480 Colorectal Cancer Cell Proliferation
In order to assess the potential use of Xn as a chemosensitizing agent able to reduce drug resistance, we studied the potential synergistic effect of Xn with SN38, a widely used chemotherapeutic drug for colorectal cancer. SW480 cells were thus pretreated for 24 h with increasing Xn concentrations. The medium was then removed and replaced with fresh medium containing increasing SN38 concentrations for 48 h before cell viability assay using crystal violet staining. Interactions between these two drugs were then analyzed by determining the combination index (CI) thanks to the Chou-Talalay equation, where CI < 1 indicates synergism, CI = 1 shows additivity, and CI > 1 antagonism. Afterwards, CI for all combination data points were calculated and plotted against Fraction Affected (FA), defined as growth inhibition potency. The normalized isobologram, characterized by different dose combinations of Xn and SN38 normalized with the dose of each drug alone, at a given x % growth inhibition, also allow to distinguish between synergism, additivity and antagonism since each dot is defined as a CI value. Most of the drug combinations show synergistic effect (green dots; CI < 1; Figure 8A), especially those reducing cell viability from 60% to 80% (0.6 < FA < 0.8) (Figure 8A), as well as the synergistic combinations plotted in normalized isobologram ( Figure 8B). Among these synergistic combinations, the one corresponding to 25 µM Xn with either 0.78 or 1.56 µM SN38 (red triangles, defined as "actual combination points", Figure 8A,B) was associated with a strong reduction of cancer cell viability (corresponding to high FA values) as compared to conventional agents alone ( Figure 8C). Thus, Xn could be used as a potential chemosensitizer to enhance anticancer properties of SN38. Since reducing the dose of synergistic combinations at high effect levels is more relevant to anticancer therapies, the dose-reduction index (DRI) of SN38 was then calculated by inverting the terms of the Chou-Talalay equation which defines the combination index. Under exposure of Xn, the dose of SN38 can be reduced (almost all DRI > 1; Figure 8D), including those of synergistic combinations that induce high effect levels (red triangles, Figure 8D). Therefore, this would allow to reduce toxicity and side effects in anticancer therapy while maintaining drug efficacy.

Discussion
In spite of chemotherapy and systematic screening for people at risk, the mortality rate of CRC has remained consistently high over the last 10 years. This low success rate in the treatment of CRC results from many failures associated with high resistance and the risk of metastasis. Therefore, in response to these therapeutic failures, new strategies have been under development for several years, aimed at increasing the effect of anticancer compounds and/or reducing their secondary effects on normal cells, thus enabling the host to better withstand chemotherapy. The present study highlights the ability of a prenylated flavonoid, Xn, to induce activation of DDR and its cellular consequences, in particular apoptosis. This induction of DNA damage by Xn constitutes a very interesting property to sensitize resistant colon cancer cells such as SW480 to classical anticancer drugs. Indeed, we found for the first time that Xn was able to cooperate positively with a metabolite of irinotecan, SN38, that could potentially enhance DNA damage in SW480 that usually exhibits a huge reduction in DNA DSB formation [10]. Synergistic combinations of Xn with SN38 leads to an increase in SW480 cell death and has the benefit of potentially decreasing the concentration of SN38, as revealed by the DRI values of SN38 (Figure 8).
The human genome is subject to permanent stress of exogenous (e.g., ionizing radiation and ultraviolet) or endogenous origin (e.g., reactive oxygen species and replication errors), which can cause lesions of varying nature, such as DNA DSB or DNA SSB. In normal cells, these lesions are detected at cell cycle checkpoints and supported by a complex network of signaling pathways commonly known as DDR. Depending on the severity of the lesions induced, DDR activates either DNA repair systems, thus allowing the resumption of cell proliferation and survival, or activates apoptosis or senescence pathways in order to maintain global genome integrity and prevent oncogenesis [42]. Genomic instability and defects in DDR (due to mutations of key molecular players) are considered as one of the characteristics of tumor cells [43]. These are not only associated with the accumulation of mutations induced in particular by the accumulation of reactive oxygen species but are also responsible for tumor cell resistance to radiotherapy and conventional anticancer drugs targeting DDR [44]. This resistance is partly linked to prolonged activation of DDR, which can reactivate DNA repair systems and therefore promote the survival, proliferation, and dissemination of malignant cells that were not initially targeted by anticancer agents [45]. The major pathways involved in DDR are the ATM and ATR pathways. ATM is specifically activated in response to DSB-type lesions, which are the most dangerous for the cell, while the ATR protein is activated in response to stress at the level of the replication forks and SSB-type lesions. Both ATM and ATR proteins phosphorylate histone H2AX (γ-H2AX) downstream, which subsequently recruits other substrates of ATM and ATR. The latter modulates the activity of key regulators of the cell cycle through activation of the transcription factor p53 and its target genes such as the p21 gene encoding the p21 protein. The latter makes it possible to trigger the G1/S, S, and G2/M checkpoints and thus to block progression of the cell cycle in order to allow DNA repair by specific enzymes, such as poly (ADP-ribose) polymerase (PARP) protein, or to induce cell death [42,46]. Very interestingly, we found that Xn was able to induce the activation of DNA damage pathways in all colon cancer cell lines but to modulate differentially the distribution of these tumor cells in the different phases of the cell cycle. The two most Xn-sensitive colon cancer cells, SW620 and HT29, both show disruption of the cell cycle in the G0/G1 and S phase, as revealed by analyzing cell distribution in phases and the key regulators of cell cycle, which are cyclins and their kinases, Cdks. We confirmed the results previously obtained by Liu et al., showing a depletion of cyclin B1 in HT29 [47], and by Logan et al., showing an arrest in the G0/G1 phase [48]. The differing results from these two authors probably result from the different concentrations used on this same HT29 line [47,48]. Here, we show that nontoxic concentrations of Xn (2.5-10 µM) were able to decrease key cyclins, such as cyclins A and B, and conversely to increase cyclin E, which controls checkpoint G1/S transition when associated with its cyclin-dependent kinases Cdk1 and 2, both of which are greatly increased depending on treatment time and concentrations (Figures 3 and 4). Very interestingly, Xn-induced cell cycle disruption is not comparable in metastatic colorectal SW620 cells, where Xn preferentially increases the SW620 cell DNA replication phase, which is again confirmed by analysis of the proteins controlling the checkpoint. Indeed, cyclin A and E protein expression is increased, whereas cyclin B is slightly decreased at the higher concentration of 10 µM ( Figure 4B). Conversely, the most resistant colon cancer cells, SW480, did no show disruption in cell cycle progression but significant apoptosis after 48 h of treatment with Xn. Besides, Xn can strongly sensitize SW480 cells to the action of a classical drug such as SN38. We demonstrated that Xn was able to activate p53, which is one of the transcriptional targets, and p21, which plays a key role in DNA damage. This is a crucial point, since the use of Xn pretreatment may not only improve the efficacy of classical chemotherapeutic drugs but also help in reducing their dose and thus in reducing their toxicity, as shown by the DRI determination of SN38 ( Figure 8D). This point is relevant, since the use of SN38 is often associated with severe gastrointestinal toxicities and neutropenia, thereby limiting its widespread use [49]. Although our combination experiments between Xn and SN38, performed in a sequential scheme (cells were pretreated for 24h Xn before the addition of SN38), displayed synergism interactions (Figure 8), the study of Ambroz et al. nicely showed in colorectal cancer cells lines (i.e., SW480, SW620, and Caco-2) that cotreatment of cells with hop-derived prenylflavonoids including Xn with chemotherapies such as 5-Fluorouracil, oxaliplatin, and irinotecan (SN38 being an irinotecan-derived potent metabolite), mostly resulted in antagonism interactions [50]. Hence, combined observations from the present study and that from Ambroz et al. suggest that synergy between Xn and chemotherapy drugs can only fully take place when drugs are sequentially administered and that simultaneous administration should be avoided because it tremendously reduces anticancer drug efficacy.

Conclusions
To date, Xn, a natural prenylated flavonoid from hops, has been described as having antioxidant [51], antiviral and antibacterial [52], antidiabetic [53], anti-inflammatory [54], and anticancer activities [55]. These properties have made it possible to classify this chalcone in the category of chemopreventive molecules. Moreover, a number of studies have been able to show the ability of Xn to prevent DNA damage induced by numerous cytotoxic agents [56][57][58]. This study highlights for the first time the ability of Xn to sensitize colon cancer cells to an anticancer agent such as SN38. This ability is associated in this case with the ability of Xn to induce lethal DNA damage and its signaling pathways. Therefore, additional studies must now be conducted with a wider panel of different anticancer drugs and cancer cell lines in order to better understand the chemosensitizing action of Xn and to investigate whether this prenylated chalcone can be classed as a potential adjuvant to increase sensitivity to anticancer agents. Funding: This work was supported by a French Government grant managed by the French National Research Agency under the program "Investissements d'Avenir", reference ANR-11-LABX-0021, and was supported by grants from the "Conseil Régional de Bourgogne" and the "Fonds Européen de Développement Régional" (FEDER).