DNA damage and necroptosis induced by peroxidase from proso millet in human colorectal cancer cells

Purpose: To investigate the effects of cationic peroxidase from proso millet (PmPOD) on DNA damage and necroptosis in human colon cancer HCT116 and HT29 cells. Methods: Cell necroptosis and cell cycle was stained using Annexin V-FITC and cell cycle kits, respectively, and evaluated by flow cytometry. Lipid raft on the membrane was disrupted by cholesterol depletion and the location of PmPOD observed by confocal microscopy. Comet assays were used to detect DNA damage, and different inhibitors were also used. Knockdown of p53 or ectopic p53 expression in HCT116 cells were transfected p53 siRNA and pCMV3-TP53-myc plasmid, and p53 expression analyzed by western blotting. Results: Pre-treatment of HCT116 and HT29 cell lines with the specific necroptosis inhibitor Nec-1 prevented PmPOD-induced necroptosis, whereas the apoptosis inhibitor, z-VAD-fmk, had no effect. The entry of PmPOD is necessary for induction of DNA damage and necroptosis. Furthermore, PmPOD induced cell cycle arrest at S phase, as well as DNA DSBs in vivo, as reflected by numerous γ-H2AX foci in CRC cells. However, the tumor suppressor protein, p53, alleviated PmPOD-induced DNA damage and necroptosis. Conclusions: These results demonstrate that PmPOD-induced DNA DSBs in CRC cells is the main cause of necroptosis, and that the tumor suppressor protein, p53, alleviates PmPOD-induced necroptosis by promoting p53-mediated repair pathways.


INTRODUCTION
Colorectal carcinoma (CRC) is the 3 rd most common malignancy worldwide, and global burden of its incidence and mortality has shifted to the developing countries in recent years, most likely driven by adoption of western-style diets [1]. Western diets have a high proportion of red meat and highly processed foods, and are deficient in vegetables and dietary fiber. These defects increase the risk for CRC. On the other hand, a healthy diet consisting largely of vegetables, fruits and whole grains, with lower intake of red meat, over processed meat and sweets, is associated with a lower risk for CRC [2,3]. The health benefits associated with the regular consumption of whole grains are most likely linked to lowering of risk factors for metabolic disorders such as type 2 diabetes and cardiovascular diseases: these risk factors are insulin resistance, dyslipidemia, inflammation and oxidative stress [4].
Necroptosis is caspase-independent programmed cell death, which has several similar morphological features with non-regulatory necrosis. It is mainly controlled by RIPK1 and RIPK3, and executed by MLKL [5,6]. Cell death is the ultimate goal of all cancer therapies, but a subset of apoptosis-resistant cancer cells may evolve. These resistant cancer cells may still be sensitive to necroptosis inducers that act via pathways independent of the apoptosis program [5]. Accordingly, cancer therapy based on necroptosis has been proposed as a novel anticancer therapeutic strategy. Cancer cells may also evade necroptosis by down-regulating regulatory molecules or by acquiring functional mutations in them. Necroptosis appears to have evolved as an alternative safeguard for inducing cell death in the event of caspase inactivation whereupon the cells cannot be eradicated by apoptosis [6].
Proso millet (Panicum miliaceum L.) is an oldest grains still cultivated and consumed in many developing countries, especially in Asia [7]. It is a highly nutritious cereal rich in protein and phytochemical compounds [7]. Studies have shown that the cationic peroxidase, PmPOD isolated from this millet induces RIPK1-and RIPK3-dependent regulatory cell necrosis in HCT116 and HT29 cells [8]. Although the exact mechanism is yet to be identified, recent studies indicate that PmPOD damages DNA through hydrolytic cleavage in vitro. The study was designed to investigate the DNA damage and necroptosis caused by PmPOD in CRC cells, and the potential mechanistic role of the tumor suppressor protein p53.

Cell culture
Human colorectal cancer cell lines (HCT116 and HT29) were gained from Chinese Center Type Culture Collection (CCTCC), and routinely cultured in 90 mm plastic tissue culture flask with RPMI Medium 1640 Basic containing 10 % FBS, 100 units/mL penicillin, and 100 μg/mL streptomycin. Two cells lines were cultured as described previously [8].

Necroptosis analysis by flow cytometric
The cell death was measured through apoptosis detection kit (C1062S, Beyotime, Shanghai, China) by flow cytometric (Becton Dickinson, Mountain View, CA). After exposure to PmPOD and/or necroptosis inhibitor Nec-1 or apoptotic inhibitor z-VAD-fmk, the HCT116 and HT29 cells were reaped by centrifugation at 600 g for 10 min at 4°C, and re-suspended cells were added to binding buffer (195 μL) provided with the kit, followed by addition of Annexin V-FITC (5 μL) and of PI (10 μL), with thorough mixing. Thereafter, the cells were kept in incubator (25 °C) for 20 min in the dark, washed thrice, and then analyzed using flow cytometer. This double staining technique was used to distinguish live cells (low level of fluorescence), early-stage apoptotic cells (Annexin + /PI ̶ ) and necrotic cells (Annexin ̶ /PI + ) from advanced-stage apoptotic cells (Annexin + /PI + ).

Cell cycle analysis
The disruption of cell cycle was analyzed using cell cycle and apoptosis analysis kit (C1052, Beyotime, Shanghai, China) according to the manual. The HCT116 and HT29 cells were seeded in 6-well plates, respectively. The number of cells was adjusted by medium, and each at a cell density of 2 × 10 5 cells/well. The both of cells were reaped and fixed in ice-cold 70 % ethyl alcohol in phosphate-buffer saline (PBS) at 4 °C for 2 h. The cells were centrifuged at 600 g for 10 min, after washing twice with PBS, the cells were incubated with the PBS staining solution containing PI and RNase A for 30 min in incubator (25°C) in the dark, and analyzed using flow cytometry. The cell cycle distribution of 10 000 cells was recorded and analyzed by FloMax software.

Lipid raft disruption by cholesterol depletion
To deplete cellular cholesterol and disrupt lipid rafts, HCT116 and HT29 cells were pretreated with 2 mM methyl--cyclodextrin (mCD) in RPMI-1640/5% FBS in incubator at 37 °C for 4 h, followed by dealing with 50 μg/mL PmPOD for 2 h. The treated cells were then transferred on ice to terminate the reaction, and the inhibition of pinocytosis was observed by confocal microscopy.

Western blotting
The HCT116 and HT29 colon cancer cells treated with different concentration PmPOD or same concentration of for specific time, and then were washed trice with pre-cooling PBS, and lysed using RIPA lysis buffer (P0013B, Beyotime, Shanghai, China). The lysates were scraped from the plates and centrifuged at 12 000 g at 4 °C for 15 min for removing cell debris. The equal amount protein samples (20 μg) were resolved using 8-12 % SDS-PAGE and then the gel was electrotransferred onto polyvinylidene difluoride (PVDF) blotting membranes (0.45 μm, A10176144, Millipore, Billerica, USA) at 0.8 mA/cm 2 for 40 min. The membranes were further blocked with blocking buffer (containing 5% skimmed milk powder and in 0.05 % Tween-20PBS) in incubator (25 °C) for 1h, and then incubated with the suitable primary antibodies (for H2AX, γH2AX PmPOD, RIPK1, RIPK3, p53, or -tubulin; diluted 1:1000 in blocking solution) and slowly agitated at 4 °C overnight. After washing thrice in TBST (Tris-HCl-buffered saline and 0.05% Tween 20) for 10 min each, the membranes were kept together with HRPconjugated secondary antibody (diluted 1:10 000 in blocking buffer) in sealed hybrid bag for 1 h at room temperature (20-25°C). After washing specific protein bands were visualized using western blotting substrate (Engreen, Beijing, China) in dark room. -Tubulin and H2AX were used as the loading control.

Neutral and alkaline comet assay
Neutral comet assay was executed as previously report [9] on HCT116 and HT29 cells after incubating cells with PmPOD with or without necroptosis inhibitors (Nec-1 and GSK-872) and ROS scavengers (NSA, BHA, BHT and NAC). Briefly, the human colon cancer cells were harvested and re-suspended in low-melting-point agarose (0.7 %, R0801, Thermo Scientific TM ) at a cell density of 10 5 cells/mL, and 100 μL of each sample was seeded on comet slides pre-coated with 1 % normal-melting-point agarose.
Coverslips were placed over the slides and the agarose was allowed to gel on a cool plate at 4 °C for about 10 min. The slides were then immersed in the lysing buffer (10 mM Tris, 2.5 M NaCl, 100 mM Na 2 EDTA, 10 % DMSO and 1 % Triton X-100, pH 10,) in the dark at 4 °C for 2 h, and washed for 10 min in TBE buffer (0.445 M Tris-HCl, 0.445 M boric acid, and also containing 0.01 M Na 2 EDTA). The slides were placed horizontally on an electrophoresis tray, which was filled with TBE solution, electrophoresed for 30 min at 21 V, and thereafter washed in 0.9 % NaCl for 2 min. The slides were neutralized using 0.4 M Tris-HCl (pH 7.5), then the they were stained with 8.0 μg/mL EB for 5 min, and observed under a fluorescence microscope with an excitation wavelength of 360 nm. The cell number of DNA comets and the percentage DNA content in the comet tail region were quantitatively measured using Kinetic Comet 4.0 software. Three independent assays and 30 cells were analyzed for each slide. For the alkaline comet assay, the slides were kept in a denaturing buffer (0.3 M NaOH, 1 M Na 2 EDTA, pH >13) for 1 h at 4 °C, and electrophoresed at 21 V for 30 min. The slides were then neutralized in 0.4 M Tris-HCl (pH 7.5) for about 5 min, followed by TBE for 2 min and finally dehydrated through an ethanol gradient of 60, 75, 90, and 100 % for 5 min at each concentration.

Immunofluorescence
For PmPOD and γ-H2AX staining, HT29 and HCT116 cells were seeded on coverslips precoated with 6-well plate, fixed with 4 % paraformaldehyde (PFA), and after rinsing twice with PBS, they were permeabilized in 1 % Triton X-100 prepared in PBS for about 10 min. The coverslips were incubated with 5 % BSA for blocking non-specific antibody binding sites inside cells at 37°C for 30 min. Then, the coverslips were incubated with anti-PmPOD, or anti-γ-H2AX (Ser139) (Bioss, Beijing, China) antibodies diluted in blocking buffer at 4 °C overnight, washed thrice, and incubated with FITC-conjugated secondary antibodies. The nuclei were counterstained with Hoechst33342. The stained samples were examined under a laser scanning confocal microscope. The wash buffer used in every step was PBS.

Statistical analysis
The distribution of all data sets was made using one-way analysis of variance (ANOVA) by using the SPSS13.0 software packages (SPSS, Inc., Chicago, IL, USA). All data are expressed as mean ± standard deviation (mean ± SD, n = 3). p < 0.05 and p < 0.01 were considered to significant difference and distinctively significant difference.

PmPOD induced colon cancer cell necroptosis
In a previous study, it was found that PmPOD time and dose dependently induced necroptosis in human colon cancer HCT116 and HT29 cells by ATP assay, and its IC 50 values were 54.87 ± 3.67 and 47.50 ± 3.89 μg/mL for the HT29 and HCT116 cells, respectively [8]. As shown in Figure 1A and B, 50 μg/mL PmPOD significantly reduced the viability of HCT116 cells, with a marked aggrandize in the necrotic or late apoptotic cells, as featured by loss of membrane integrity which made for easy staining with PI. After 24 h of treatment, the necrotic cells increased to 32.41 % in HCT116 cells and 26.11 % in HT29 cells, when compared to 5.23 and 2.23 % in the respective untreated controls. However, no significant increase was seen in the proportion of Annexin V + /PI ̶ (early apoptotic cells) following PmPOD treatment. In addition, reduction of cell viability triggered by PmPOD was not blocked by the apoptotic inhibitor z-VADfmk, which rather aggravated the toxic effects of PmPOD (Figures 2A and B). Following pretreatment with the necroptosis inhibitor Nec-1(RIPK1 inhibitor, CAS: 4311-88-0), PmPODinduced cell death was significantly attenuated (Figures 2A and B). These finding are consistent with previous reports, indicate that PmPOD induced necroptosis in CRC cells [8].

Cholesterol reduction inhibited PmPODinduced cell death
Incubation of HCT116 and HT29 cells with 2 mM The viability of HCT116 cells following treatment with 50 μg/mL PmPOD for 2 and 24 h were 89.12 and 61.1 %, respectively, while the corresponding viabilities in the HT29 cells under the same treatment for 2 and 24 h were 83.28 and 52.1 %, respectively. However, when cotreated with mCD, the viability values after 2 and 24 h increased to 95.01 and 87.34 %, respectively in HCT116 cells, and 95.9 and 86.2 %, respectively in HT29 cells. These results indicate that cholesterol depletion inhibited PmPOD endocytosis in CRC cells, thereby increasing their viability.

PmPOD induced cell cycle arrest at S phase
To determine the potential effect of PmPOD on the cell cycle dynamics of HCT116 and HT29 The number of HT29 cells at S phase increased from 29.41 to 35.66, 38.76 and 39.36 % following treatment with 10, 25 and 50 μg/mL PmPOD, respectively ( Figure 4 C and D). The cells at the "sub-G1 area" represent apoptotic cells in the sample. In the analysis of PmPOD-treated cells, the number of cells in the sub-G1 phase was small and negligible. In HCT116 cells treated with 10, 25 and 50 μg/mL PmPOD, the percentages of sub-G1 phase cells were 0.10, 0.12 and 0.59 %, respectively, while that of the untreated control group was 0.08 %. In contrast, in HT29 cells treated with 10, 25 and 50 μg/mL PmPOD, the percentages of sub-G1 phase cells were 0.12, 0.54 and 0.08%, respectively, while that of the untreated control group was 0.06%. These consequences suggest that the PmPODmediated reduction in CRC cell viability is associated with cycle arrest at the S phase.

DNA DSBs induced by PmPOD in CRC cells
Alkaline and neutral comet assays are sensitive and rapid technique for evaluate various forms of DNA damage and DNA DSBs, respectively. As is shown in Figure 5A and B, majority of the HCT116 and HT29 cells treated with 25 or 50 μg/mL PmPOD for 24 h had long comet tails in both assays, in contrast to the untreated controls, indicating that PmPOD induced overall DNA damage as well as DSBs. In addition, quantification of the alkaline and neutral comet assays revealed that PmPOD treatment not only increased the number of cells with DNA damage, but also the DNA content within the tails ( Figure  5C and D). Furthermore, the DNA content in the comet tails of HT29 cells were higher than that of the HCT116 cells (37 % vs 25 %).

PmPOD induced formation of γ-H2AX foci
Rapid phosphorylation of H2AX at Ser-139 of the minor histone H2A variant to produce γ-H2AX, is an early cellular response to DSBs, which is making it a suitable marker of DNA DSBs [10]. The number of γ-H2AX foci per nucleus increased significantly in both HCT116 and HT29 cells incubated for 6 h with 25 or 50 μg/mL PmPOD ( Figure 6A). Interestingly, most of the foci were clustered around the nuclei rather than within them in the HT29 cells (Figure 6 B). These in situ immunofluorescence findings were validated by western blotting of the total cellular proteins (Figure 6 C). Furthermore, γ-H2AX levels increased significantly in a time-dependent manner following 25 μg/mL PmPOD treatment for 6, 12 and 24 h (Figure 6 D). In contrast, a sharp increase in γ-H2AX was seen after 12 h of incubation with 50 μg/mL PmPOD, but no further increase occurred at 24 h (Figure 6 D). These results indicate that the generation of DNA DSBs is a committed step that precedes PmPODinduced necroptosis of CRC cells.

PmPOD-induced DNA damage and necroptosis were associated with p53 expression
To determine the potential role of p53 expression in the cellular response to PmPOD, p53 expression was silenced in HCT116 cells before PmPOD treatment (Figure 7 B). The p53 knockdown markedly altered nuclear morphology and aggravated DNA damage in the PmPODtreated cells (Figure 7 A), and significantly enhanced PmPOD-induced necroptosis (Figure 7 C). In addition, ectopic p53 expression via the pCMV3-TP53-myc plasmid (Figure 7 D) significantly reduced the DNA damage triggered by PmPOD (Figure 7 A), and de-sensitized the HCT116 cells to PmPOD-induced cell death (Figure 7 E). It was also found that PmPOD induced elevation in ROS in colon cancer HCT116 and HT29 cells, and that RIPK1 and RIPK3 conduced to PmPOD-induced ROS overproduction ( Figure 8). Both of HCT116 and HT29 cells treated with PmPOD (50 g/mL) had longer comet tails. The appearance of the comet tails in cells was not significantly suppressed in the presence of necroptosis inhibitors (Nec-1, GSK-872, and NSA), nor was it significantly prevented in the presence of ROS scavengers (BHA, BHT or NAC) (Figure 9). These findings indicate that the DNA DSBs of PmPOD on the CRC cells is dependent on p53 expression.

DISCUSSION
This study has demonstrated that PmPODtriggered DNA DSBs in human colon cancer HCT116 and HT29 cells are the main cause of necroptosis. DNA DSBs are potentially lethal DNA lesion, which may arise spontaneously during DNA replication or consequences of external disadvantage, such as ionizing radiation (IR), oxidative stress and chemotherapeutic drugs [11,12]. DNA DSBs can be lethal if not repaired or if it is mis-repaired. Therefore, DSBs sensors induce cell cycle arrest at the S phase, in which allow the cells to undergo DNA repair. It was found that PmPOD induced cell cycle arrest at the S phase in CRC cells.
Cells undergo DNA replication during S phase, and since accurate replication is necessary to prevent genetic aberrations which can lead to apoptosis or other pathological conditions, this stage of the cell cycle are crucial for detecting and repairing any DNA damage. When the replication fork arrives at the damaged site, protein kinase is activated [13], which in turn triggers ATM-and ATR-mediated DNA damage response [11]. Both pathways activate the cellcycle-checkpoints, which either allow the cells to repair the damage or prevent the proliferation of the damaged cells by inducing senescence, or in case of extensive damage, may also lead to apoptotic cell death.
An important step in ATM-and ATR-dependent signaling is the phosphorylation of H2AX at Ser-139 to produce γ-H2AX, which accumulates at the DSBs sites and is thus a suitable and sensitive molecular maker of DNA damage [12]. In addition to recruiting DNA repair proteins, γ-H2AX also recruits the apoptosis inducing factor (AIF) from the mitochondria to form DNAdegrading complex during necroptosis. Therefore, production DNA DSBs may be a crucial cause leading to cells undergoing necroptosis [12,14].
The p53 is not only tumor-suppressing gene, but also a stress-responsive transcription factor that can activate different target genes in response to diverse stimuli, including oxidative stress, genotoxic damage and hypoxia [15]. Therefore, it is also known as "the guardian of the genome" for its role in preserving the stability of the genome. Mutation p53 or loss of normal p53 function is strongly associated with an increased susceptibility to various types of cancer [16]. P53 protects cells against malignant transformation, and is deleted or mutated in approximately 50 % of human cancers [13].
In the present study, ectopic expression of p53 significantly alleviated the PmPOD-induced DNA damage in human colon cancer cells. The DNA damage pathway stabilizes p53, which then transcriptionally activates a number of genes regulating cell cycle arrest or apoptosis [17], and hence cell fate [18]. Amongst the p53-target genes, p21 induces cell cycle arrest at G 1 and S phase, which provids time for DNA repair and promotes cell survival [18] [19]. In contrast, Bax, PUMA and Noxa are crucial for p53-dependent apoptosis [20][21][22]. Recent studies have revealed that inhibition of intracellular ROS (induced by shikonin) with scavenger suppressed DNA DSBs caused by H 2 O 2 , suggesting that ROS also play a very important role in regulation of DNA DSBs [12,23]. Basis on experimental results of this study, it can be hypothesized that p53 plays an important role in repairing PmPOD-induced DNA damage and necroptosis. This may be due to the selective activation of p53-target genes. However, this needs to be further validated.
Fluid-phase endocytosis or pinocytosis, which is a non-receptor-mediated endocytosis, refers to the uptake of macromolecules from the extracellular fluid into cytoplasm by membranebound endocytotic vesicles [24]. mCD suppresses fluid-phase endocytosis by removing cholesterol from the plasma membrane [25]. Horseradish peroxidase (HRP) is a representative of class III heme peroxidases [26]. Cerebral endothelial cells take up HRP by pinocytosis, a non-receptor-mediated endocytosis, and the protein is stable inside the cell for at least several hours [27]. Transient expression of HRP in mammalian cells does not produce serious toxicity [28,29]. However, sustained expression of HRP in cells brings about morphological changes and death, although the underlying mechanism is still unknown [30]. Being a cationic peroxidase, PmPOD-induced tumor cell necroptosis is a function of the mode of entry of PmPOD into the cells. To determine the mechanism of PmPOD internalization, its endocytosis in mCD-treated and contrastive cells were tracked via immunofluorescence. In this study, it was found that PmPOD was taken up by fluid-phase endocytosis, since its intracellular uptake was disrupted by mCD-mediated cholesterol depletion, which also inhibited cell necroptosis. These results indicate that PmPOD entry is necessary for induction of DNA damage and necroptosis.

CONCLUSION
This paper demonstrates that DNA DSBs induced by PmPOD in CRC cells is the main cause of necroptosis, and that the tumor suppressor protein p53 alleviates PmPODinduced necroptosis by promoting cell survival through p53-mediated repair pathways.