Intestinal commensals are potential important contributors to the etiology of sporadic colorectal cancer, but mechanisms by which bacteria can initiate tumors remain uncertain. Herein, we describe mechanisms that link Enterococcus faecalis, a bacterium known to produce extracellular superoxide, to the acute induction of chromosomal instability. Immortalized human and nontransformed murine colonic epithelial cells, along with a mouse colonic ligation model, were used to assess the effect of E. faecalis on genomic DNA stability and damage. We found that this human intestinal commensal generated aneuploidy, tetraploidy, and γH2AX foci in HCT116, RKO, and YAMC cells. In addition, direct exposure of E. faecalis to these cells induced a G2 cell cycle arrest. Similar observations were noted by exposing cells to E. faecalis–infected macrophages in a dual-chamber coculture system for detecting bystander effects. Manganese superoxide dismutase, catalase, and tocopherols attenuated, and caffeine and inhibitors of glutathione synthase exacerbated, the aneugenic effects and linked the redox-active phenotype of this intestinal commensal to potentially transforming events. These findings provide novel insights into mechanisms by which E. faecalis and intestinal commensals can contribute to cellular transformation and tumorigenesis. [Cancer Res 2008;68(23):9909–17]

Intestinal commensals are potential, important contributors to the development of sporadic human colorectal cancer (CRC), but the identification of relevant microorganisms has been impeded by the complex ecology of the large intestine (1, 2). Several bacteria are definitively linked to gastrointestinal tumorigenesis; for example, Helicobacter pylori is considered a class I carcinogen for gastric cancers (3). However, in contrast to the stomach, wherein the microbiota only consists of Helicobacter spp., the human colon contains >400 species with >1011 colony-forming units (cfu) per gram of feces (4). At least 75% of these microorganisms cannot be cultivated using current techniques and, therefore, remain largely uncharacterized. These difficulties have made the epidemiologic approach to identifying CRC-associated microorganisms problematic. An alternate approach involves focusing on commensals that can confer long-term exposure to the host and are potentially mutagenic.

Enterococcus faecalis is a minority constituent of the intestinal microbiota that uniquely produces reactive oxygen species (ROS), including extracellular superoxide (5). This phenotype is rare among bacteria and is known to damage colonic epithelial DNA (6). E. faecalis also causes inflammation and CRC in interleukin-10 knockout mice (7, 8) and promotes chromosomal instability (CIN) in mammalian cells (9). CIN is the most common form of genomic instability in solid tumors. It consists of gains and losses of chromosome segments or whole chromosomes, along with inversions, translocations, and complex rearrangements in the genome sequence and structure (10). The dramatic and unstable changes in genomic content for cells with CIN are typified by aneuploidy. Recent mathematical modeling and experimental evidence suggest that CIN is an early event in tumorigenesis (11, 12). The origin of CIN, however, remains uncertain. We recently showed that E. faecalis promotes CIN through free radical mechanisms initiated by superoxide (9). This effect was linked to the expression of cyclooxygenase-2 (COX-2) in macrophages and production of clastogens (or chromosome-breaking factors) that diffuse into neighboring cells to damage DNA.

The production of clastogens by E. faecalis–infected macrophages is analogous to the induction of CIN through the radiation-induced bystander effect (13). This effect is generated when cells are activated by radiation to produce diffusible mediators that damage DNA in neighboring, unirradiated cells. The bystander effect has been shown to occur in vivo in congenic sex-mismatch bone marrow transplantation experiments (14). Although causative mediators are not characterized, the bystander effect is associated with COX-2 and DNA-damaging products from this inflammatory pathway may be important (9, 15, 16).

A macrophage-induced bystander effect is consistent with the potential contribution of innate immune cells to tumorigenesis through prostaglandins, cytokines, and chemokines that have transforming, proliferative, and metastatic consequences (17). In this study, we found that E. faecalis induced aneuploidy, tetraploidy, and DNA damage in colonic epithelial cells. Superoxide production was associated with DNA double-strand breaks (DSB), G2 arrest, and changes in chromosome number. These findings occurred through a bystander effect and provide a mechanism by which redox-active intestinal commensals may contribute to cellular transformation and tumorigenesis.

Cell lines and bacterial strains. HCT116 and RKO cells are near-diploid human colonic epithelial cell lines (43–45 and 46–49 chromosomes per cell, respectively) with defective mismatch repair (American Type Culture Collection (ATCC); ref. 18). These cell lines were maintained in McCoy's 5A (Invitrogen) containing 10% fetal bovine serum (FBS), penicillin G, and streptomycin and have stable karyotypes (18, 19). YAMC cells are a nontransformed line from the Ludwig Institute for Cancer Research and were grown in 5% CO2 at 33°C using RPMI 1640 (Invitrogen) supplemented with 5% FBS, recombinant murine IFN-γ (Peprotech), and ITS premix (20). RAW264.7 cells (ATCC) are murine macrophages and were maintained as previously described (9).

E. faecalis OG1RF is a human isolate that produces extracellular superoxide (21). Among healthy adults, E. faecalis varies in concentration in feces from <105 to 108 cfu g−1 (22). This strain was grown in brain-heart infusion (Difco) at 37°C and washed in PBS with the multiplicity of infection (MOI) defined as bacterial cfu per cultured cell. Superoxide production was verified using ferricytochrome c reduction (5). MnSOD was purchased from Roche, and additional reagents were from Sigma, unless otherwise specified. Tocopherols (Tama Chemicals) and γ-carboxyethyl hydroxychroman (Encore Pharmaceuticals, Inc.) were provided by Kenneth Hensley.

For cells directly exposed to E. faecalis, a 2-h treatment was carried out in a medium without serum or antibiotics and restored to a complete medium after treatment. The pH of the medium during experiments was maintained in a physiologic range during coculture with E. faecalis. Antioxidants and inhibitors were added with E. faecalis in experiments, as specified. Experiments with the dual-chamber coculture system were carried out, as previously described (9).

Cytogenetics. Anaphase bridging was analyzed using cells grown on CultureSlides (BD Biosciences). After 24 h incubation, cells were fixed with 2% paraformaldehyde and stained with propidium iodide. Anaphases were scored by one of us (X.W.) using fluorescent microscopy according to previously described criteria (23).

Fluorescence-activated cell sorting analysis for aneuploidy and tetraploidy. Ploidy was determined by fluorescence-activated cell sorting (FACS) analysis, as previously described (24). In brief, cells were treated with 1 × 109 cfu mL−1E. faecalis for 2 h. After 48 h, cells were arrested in mitosis with colcemid, fixed (70% ethanol), and stained with phosphorylated (p-) histone H3 (Ser10) mouse monoclonal antibody (1:400 dilution, Cell Signaling Technology). Phosphorylation of histone H3 is a marker of chromosome condensation and mitosis. Cells were secondarily stained with the Alexa Fluor 488 F(ab')2 fragment of goat anti-mouse IgG (H+L; 1:1,000 dilution, Invitrogen) and propidium iodine. Cell sorting was performed on FACSCalibur (BD Bioscience) using CellQuest software.

To investigate the induction of aneuploidy, E. faecalis–exposed HCT116 cells were double-stained with Hoechst 33342 and propidium iodide. Cells with DNA content of >5N were sorted as single cells into wells onto 96-well plates (inFlux Cell Sorter, Cytopeia) and incubated for 2 wk. Colonies were enumerated and compared with untreated controls.

Cell cycle assay. Cells were synchronized in G1 using 2 mmol/L thymidine, followed by thymidine-free medium, and then thymidine-supplemented medium (25). Cells were treated with 1 × 109 cfu mL-1E. faecalis for 2 h (MOI = 1,000) and fixed. For the dual-chamber system, uninfected and E. faecalis–infected macrophages were placed in the upper compartment (9). E. faecalis is readily phagocytosized by macrophages and survives as an intracellular pathogen (26). Synchronized cells in the lower compartment were used as targets, fixed, stained with propidium iodine, and sorted by FACS. Data were analyzed using ModFit LT (Verity Software House).

Western blotting. Protein extraction and blotting were carried out using enhanced chemiluminescence (GE Healthcare). Phosphorylated ataxia telangiectasia–mutated (ATM; Ser1981), p-ATM and Rad3-related (ATR; Ser428), p-Chk1 (Ser345), p-Chk2 (Ser33/35), and p-Cdk1 (Tyr15) monoclonal antibodies and polyclonal anti–cyclin B1 antibody were purchased from Cell Signaling Technology. Murine anti–β-actin loading control was purchased from BioVision. Donkey anti-mouse IgG–horseradish peroxidase (HRP) conjugate (Santa Cruz Biotechnology) and goat anti-rabbit IgG-HRP conjugate (Zymed) were used as secondary antibodies.

In vivo DSBs caused by E. faecalis. To assess the DNA-damaging effect of E. faecalis in vivo, we developed a colonic ligation model (27). Colons of wild-type BALB/c mice were surgically filled with 1.0 mL of PBS as a control (n = 3) or E. faecalis at 1 × 108 cfu mL-1 in PBS (n = 3), and 5 mmol/L d-glucose were added to initiate superoxide production (5). Colon contents were retained using a rectal suture and ligature at the proximal colon. Care was taken to preserve blood flow with no ischemia observed in the colons of 24 mice previously studied using this protocol (27). At 6 h postinoculation, colons were removed for analysis.

Wild-type C57BL/6 mice were subjected to 13 Gy whole body radiation in a 137Cs irradiator. Animals were euthanized 4.5 h after irradiation, and colons were removed for analysis. Protocols were approved by animal study committees of the University of Oklahoma Health Sciences Center and Oklahoma City VA Medical Center.

Immunofluorescence and immunohistochemistry. HCT116 and RKO cells treated with E. faecalis were fixed with 2% paraformaldehyde and stained using p-histone H2AX (Ser139) rabbit monoclonal antibody conjugated to Alexa Fluor 488 (Cell Signaling). DNA was counterstained with 4′-6-diamidino-2-phenylindole. γH2AX-positive cells were identified by laser scanning confocal microscopy (Leica Microsystems) or enumerated by FACSCalibur and analyzed using Summit software (v4.3, DAKO).

Epitope retrieval of paraffin-embedded colon sections was done using a pressurized decloaking chamber (Biocare Medical, LLC). Sections were blocked using avidin/biotin blocking kit (Vector Lab), and peroxidase activity was quenched using 3% H2O2. After washing, slides were incubated in horse serum and bovine albumin and exposed to rabbit anti–phospho-H2AX antibody (Cell Signaling). Slides were incubated in biotinylated donkey anti-rabbit secondary antibody (Jackson Immuno Research Lab) and SA-HRP (Dako), and chromogenic development was performed using 3,3′-diaminobenzidine substrate (brown) with hematoxylin as the counterstain (Biocare Medical). Slides were scored for γH2AX in a blinded fashion by one of us (S.L.), counting only nuclear staining as positive.

Statistical analysis. Data were expressed as means with SDs. Experimental and control groups in ploidy assays were compared by χ2 test and the number of positive nuclei in colon biopsies by the Student's t test. P values of <0.05 were considered statistically significant.

E. faecalis promotes anaphase bridging, aneuploidy, and tetraploidy. To assess mechanisms by which E. faecalis damages DNA and causes CIN, we initially evaluated the diploid and chromosomally stable HCT116 cell line for anaphase bridging after exposure to E. faecalis. Anaphase bridges appear as extended chromosomes between spindle poles during the separation of sister chromatids (23, 28). HCT116 cells showed numerous abnormal mitotic patterns after exposure to E. faecalis, including lagging chromosomes, anaphase bridges, and tripolar anaphases (Fig. 1A). We found a significant increase in the frequency of anaphase bridging after E. faecalis exposure (86 of 151 anaphases, 57%) compared with untreated control cells (35 of 194 anaphases, 18%; P < 0.001). The frequency of anaphase bridging was significantly reduced when MnSOD was added (48 of 158, 30%; P < 0.001), suggesting that superoxide promoted abnormal mitoses in these cells.

Figure 1.

E. faecalis causes anaphase bridging, aneuploidy, and tetraploidy. A, HCT116 cells after exposure to E. faecalis: normal anaphase (i), abnormal mitosis (ii), anaphases with lagging chromosomes (iii–v), and tripolar mitosis with bridging (vi). B, histograms of FACS analysis for HCT116 cells: untreated cells (left), cells exposed to E. faecalis for 2 h (middle), and cells exposed to E. faecalis for 2 h with addition of MnSOD (right). p-Histone H3 antibody (Y axis) with mitotic cells separated from interphase cells and propidium iodine for DNA content (X axis). R2 gate contains cells with DNA between 2N and 4N (hypodiploid); R3 gate contains 4N DNA; R4 gate contains DNA between 4N and 8N; R5 gate contains >8N DNA. R2 + R4 + R5 represent the proportion of aneuploid and tetraploid cells. C, the proportion of aneuploid and tetraploid cells increases from 24 to 96 h after E. faecalis treatment (gray columns) compared with untreated control (hashed columns). MnSOD protects against aneuploidy and tetraploidy at 48 and 72 h (white columns). D, α-tocopherol, γ-tocopherol, and γ-carboxyethyl hydroxychroman (γ-CEHC) decrease the frequency of aneuploidy and tetraploidy at 48 h compared with E. faecalis alone. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 1.

E. faecalis causes anaphase bridging, aneuploidy, and tetraploidy. A, HCT116 cells after exposure to E. faecalis: normal anaphase (i), abnormal mitosis (ii), anaphases with lagging chromosomes (iii–v), and tripolar mitosis with bridging (vi). B, histograms of FACS analysis for HCT116 cells: untreated cells (left), cells exposed to E. faecalis for 2 h (middle), and cells exposed to E. faecalis for 2 h with addition of MnSOD (right). p-Histone H3 antibody (Y axis) with mitotic cells separated from interphase cells and propidium iodine for DNA content (X axis). R2 gate contains cells with DNA between 2N and 4N (hypodiploid); R3 gate contains 4N DNA; R4 gate contains DNA between 4N and 8N; R5 gate contains >8N DNA. R2 + R4 + R5 represent the proportion of aneuploid and tetraploid cells. C, the proportion of aneuploid and tetraploid cells increases from 24 to 96 h after E. faecalis treatment (gray columns) compared with untreated control (hashed columns). MnSOD protects against aneuploidy and tetraploidy at 48 and 72 h (white columns). D, α-tocopherol, γ-tocopherol, and γ-carboxyethyl hydroxychroman (γ-CEHC) decrease the frequency of aneuploidy and tetraploidy at 48 h compared with E. faecalis alone. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Close modal

Because anaphase bridging can lead to aneuploidy (23), we next determined whether exposure to E. faecalis resulted in aneuploidy. Mitotic HCT116 cells were sorted using a monoclonal antibody to p-histone H3 and DNA content determined by FACS. Approximately, 10% of untreated mitotic HCT116 cells were nondiploid with a DNA content greater or less than 4N (Fig. 1B,, left). After a 2-hour exposure to E. faecalis, the average frequency of nondiploid cells was significantly increased at 24, 48, 72, and 96 hours postexposure (Fig. 1B,, middle, and C). This correlated with significant increases in both aneuploid and tetraploid cells (i.e.,, 8N DNA content for cells in G2 or mitosis; 8.3% compared with 0.6% for untreated control; P < 0.001). Similarly, the percentage of cells in G2 (i.e.,, nonmitotic) with 8N DNA content increased after E. faecalis treatment compared with untreated controls (7.8% versus 0.0%; P < 0.001). MnSOD was associated with a reduction in the number of mitotic aneuploid cells at 24, 48, and 72 hours (Fig. 1B,, right, and C). Reductions in mitotic cells and cells in G2 with 8N DNA content were also observed (8.2 to 3.5% and 7.8 to 2.0%, respectively, at 48 hours). We next assessed whether tocopherols, scavengers of lipid radicals, protected against aneuploidy and tetraploidy. Addition of α-tocopherol or γ-tocopherol significantly decreased the percentage of aneuploid cells compared with treatment with E. faecalis alone (Fig. 1D). Similar reductions were seen in the percentage of tetraploid cells. Notably, the tocopherol metabolite, γ-carboxyethyl hydroxychroman, a COX-2 inhibitor but poor scavenger (29), also decreased the frequency of aneuploidy (Fig. 1D).

To extend these findings, we repeated experiments using RKO cells. The percentage of aneuploid and tetraploid cells at baseline was similar to that for HCT116 cells and increased significantly after exposure to E. faecalis (Supplementary Fig. S1). In addition, the percentage of nondiploid cells decreased when MnSOD was added, suggesting that superoxide from E. faecalis promoted changes in chromosome number for both cell lines.

Errors in mitotic segregation, e.g., due to anaphase bridging or mitotic failure, can result in aneuploidy and tetraploidy (10, 30). To determine whether E. faecalis generated long-term changes in chromosome number, we exposed HCT116 cells to E. faecalis and selected highly aneuploid or tetraploid cells by FACS (i.e., DNA content of >5N). Single cells were sorted, and after 2 weeks, only 20 of 288 cells (6.9%) grew compared with 51 of 288 untreated cells (17.7%; P < 0.001). Growth of untreated nondiploid cells was similar to untreated diploid cells with 60 of 288 cells (20.8%) viable (P = 0.40). E. faecalis–exposed clones were then somatically selected over 20 generations for heritable changes in DNA content. Reexamination by FACS detected only one tetraploid clone (Supplementary Fig. S2, right), with the remainder being diploid (Supplementary Fig. S2, middle). This implied that the acute induction of aneuploidy and tetraploidy by E. faecalis was unstable (or lethal) with few clones able to propogate only by reverting back to diploidy or, on occasion, maintaining tetraploidy.

E. faecalis activates DNA damage responses and G2 arrest. We noted that HCT116 and RKO cells developed megalocytosis after exposure to E. faecalis (data not shown). This effect was abrogated by catalase, but not MnSOD, and seemed similar to cytopathic effects produced by bacterial cyclomodulins known to interfere with the eukaryotic cell cycle. Cyclomodulins, such as the cytolethal distending toxin, delay entry into mitosis by damaging host cell DNA (25). We, therefore, analyzed by FACS HCT116 and RKO cells after exposure to E. faecalis and found that both developed arrest at the G2-M transition (Fig. 2A).

Figure 2.

E. faecalis activates DNA damage responses. A, cell cycle analysis shows E. faecalis induced a G2-M cell cycle arrest in HCT116 cells (left) and RKO cells (right). Top, untreated control; bottom, cells treated with 1 × 109 cfu mL-1E. faecalis for 2 h. Syn, cells synchronized to G1 by double thymidine block. B and C, Western blots of p-ATM (Ser1981), p-ATR (Ser428), p-Chk1 (Ser345), p-Chk2 (Ser33/35), p-Cdk1 (Tyr15), and cyclin B1. ATM/ATR is activated by transient infection with E. faecalis in HCT116 cells (B) and RKO cells (C). Dephosphorylation of Cdk1 is evident at 48 h in both cell lines. Phosphorylation of Cdk1 at Tyr15 was seen in untreated cells for both cell lines (bottom). D, caffeine blocks ATM/ATR leading to increased frequency of aneuploidy and tetraploidy in E. faecalis–exposed HCT116 and RKO cells. Black columns, E. faecalis; white columns, E. faecalis with 2.5 mmol/L caffeine. ***, P < 0.001 compared with E. faecalis alone.

Figure 2.

E. faecalis activates DNA damage responses. A, cell cycle analysis shows E. faecalis induced a G2-M cell cycle arrest in HCT116 cells (left) and RKO cells (right). Top, untreated control; bottom, cells treated with 1 × 109 cfu mL-1E. faecalis for 2 h. Syn, cells synchronized to G1 by double thymidine block. B and C, Western blots of p-ATM (Ser1981), p-ATR (Ser428), p-Chk1 (Ser345), p-Chk2 (Ser33/35), p-Cdk1 (Tyr15), and cyclin B1. ATM/ATR is activated by transient infection with E. faecalis in HCT116 cells (B) and RKO cells (C). Dephosphorylation of Cdk1 is evident at 48 h in both cell lines. Phosphorylation of Cdk1 at Tyr15 was seen in untreated cells for both cell lines (bottom). D, caffeine blocks ATM/ATR leading to increased frequency of aneuploidy and tetraploidy in E. faecalis–exposed HCT116 and RKO cells. Black columns, E. faecalis; white columns, E. faecalis with 2.5 mmol/L caffeine. ***, P < 0.001 compared with E. faecalis alone.

Close modal

We hypothesized that E. faecalis-induced G2-M arrest was due to DNA damage from ROS and, therefore, investigated phosphorylations involved in checkpoint activation. ATM and ATR are kinases that initiate DNA damage response signal transduction cascades (31). After DNA damage, ATM and/or ATR are activated by autophosphorylation and, in turn, phosphorylate Chk1 and/or Chk2 (3234). Chk1/Chk2 further phosphorylates Cdc25C leading to sequestration in the cytoplasm and failure to activate Cdk1 (cyclin B–dependent kinase 1), a key regulator of mitotic progression. Phosphorylation of Cdk1 at Thr161 and dephosphorylation at Tyr15/Thr14 activates Cdk1, leading to progression from G2 to M (35). Western blots for p-ATM/p-Chk2 and p-ATR/p-Chk1 showed activation in HCT116 and RKO cells after exposure to E. faecalis (Fig. 2B and C). However, unlike HCT116 cells, RKO cells expressed p-ATR at baseline, suggesting constitutive activation, as has been reported for many colon cancers (36). Despite the G2-M arrest after exposure to E. faecalis (Fig. 2A), phosphorylation of Cdk1 at Tyr15 failed to increase (Fig. 2B and C), although doxorubicin, the positive control, produced this effect. Cyclin B1 levels, in contrast, were markedly increased in treated cells compared with controls.

To investigate whether the spindle-assembly checkpoint contributed to E. faecalis-induced cell cycle arrest, we determined the proportion of cells containing p-histone H3 as a marker of chromosome condensation and mitosis. There was, however, no change in the proportion of cells with p-histone H3 after exposure to E. faecalis (Supplementary Fig. S3A and B). Nocodazole-treated cells, in contrast, showed a 5-fold increase by 24 hours. This provided evidence that the spindle-assembly checkpoint was not activated by E. faecalis. Finally, caffeine, an ATM/ATR inhibitor (37), significantly increased the percentage of aneuploid and tetraploid cells after exposure to E. faecalis (Fig. 2D), and partially rescued cells from E. faecalis–induced G2 arrest (Supplementary Fig. S3C). In sum, these findings indicated that E. faecalis–induced G2 arrest did not occur through the canonical Cdk1 pathway, and the elucidation of underlying mechanisms are under investigation.

E. faecalis forms γH2AX foci. Cellular responses to DNA damage include DNA repair cascades and, for severe injury, apoptosis. In mammalian cells, γH2AX is rapidly formed by ATM at sites of DSBs. We identified γH2AX foci in the nuclei of cells after treatment with 1 μmol/L doxorubicin, an inducer of DSBs, and after exposure to E. faecalis (Fig. 3A and Supplementary Fig. S4). The proportion of γH2AX-positive cells was bimodal with peaks appearing immediately after exposure to E. faecalis and 48 hours later (Fig. 3B). The proportion of γH2AX-positive cells was also noted to increase with greater MOIs (Fig. 3C). Finally, caffeine decreased the percentage (±SD) of cells positive for γH2AX (E. faecalis versus E. faecalis with 2.5 mmol/L caffeine: 43.4 ± 0.5 versus 20.2 ± 1.8 for HCT116 cells and 79.2 ± 0.2 versus 67.4 ± 0.4 for RKO cells; P < 0.01 for both comparisons). These observations show that E. faecalis generates γH2AX foci and inhibition of ATM/ATR promotes E. faecalis–induced aneugenesis.

Figure 3.

E. faecalis causes DSBs in colonic epithelial cells. A, laser scanning confocal microscopy of HCT116 cells exposed to E. faecalis at an MOI of 1,000 shows DSBs as γH2AX foci (green nuclei). DNA was counterstained using 4′,6-diamidino-2-phenylindole (blue nuclei) with images merged (right). No γH2AX foci occur in untreated controls (top). B, analysis of γH2AX-positive cells by FACS shows DSBs after exposure to E. faecalis with a bimodal effect: blue trace, HCT116 cells; red trace, RKO cells. RKO cells are more sensitive to E. faecalis–induced DSBs at 0 and 6 h compared with HCT116 cells, but not beyond 24 h. C, E. faecalis causes DSBs in a dose-dependent manner for HCT116 cells (top) and RKO cells (bottom). Red trace, untreated control; gray trace, cells exposed to E. faecalis at a MOI of 500; green trace, MOI of 1,000. D, MnSOD is partially protective against DSBs in HCT116 cells at 0 h (top left), but not 24 h (top right) after exposure to E. faecalis. In contrast, MnSOD has no effect on the proportion of γH2AX-positive RKO cells at 0 h (bottom left) and 24 h (bottom right). Catalase eliminates DSBs for HCT116 and RKO cells at all time points: gray trace, untreated control; red trace, MOI of 1,000; green trace, addition of MnSOD (1,200 units mL-1); blue trace, addition of catalase (1,200 units mL-1).

Figure 3.

E. faecalis causes DSBs in colonic epithelial cells. A, laser scanning confocal microscopy of HCT116 cells exposed to E. faecalis at an MOI of 1,000 shows DSBs as γH2AX foci (green nuclei). DNA was counterstained using 4′,6-diamidino-2-phenylindole (blue nuclei) with images merged (right). No γH2AX foci occur in untreated controls (top). B, analysis of γH2AX-positive cells by FACS shows DSBs after exposure to E. faecalis with a bimodal effect: blue trace, HCT116 cells; red trace, RKO cells. RKO cells are more sensitive to E. faecalis–induced DSBs at 0 and 6 h compared with HCT116 cells, but not beyond 24 h. C, E. faecalis causes DSBs in a dose-dependent manner for HCT116 cells (top) and RKO cells (bottom). Red trace, untreated control; gray trace, cells exposed to E. faecalis at a MOI of 500; green trace, MOI of 1,000. D, MnSOD is partially protective against DSBs in HCT116 cells at 0 h (top left), but not 24 h (top right) after exposure to E. faecalis. In contrast, MnSOD has no effect on the proportion of γH2AX-positive RKO cells at 0 h (bottom left) and 24 h (bottom right). Catalase eliminates DSBs for HCT116 and RKO cells at all time points: gray trace, untreated control; red trace, MOI of 1,000; green trace, addition of MnSOD (1,200 units mL-1); blue trace, addition of catalase (1,200 units mL-1).

Close modal

ROS promote γH2AX foci. The temporal pattern of γH2AX formation (Fig. 3B) suggested rapid DNA damage (38). To confirm that ROS generated γH2AX foci and, presumably, DSBs, MnSOD, and catalase were added to cells exposed to E. faecalis. Compared with E. faecalis alone, MnSOD partially protected against γH2AX formation in HCT116 cells at 0 hour, but not 24 hours (Fig. 3D,, top), and failed to protect RKO cells at any time point (Fig. 3D , bottom). Catalase, however, prevented the formation of γH2AX foci, suggesting that the spontaneous dismutation of superoxide to H2O2 generated most of the DNA damage in these cells.

E. faecalis induces DNA damage and G2 arrest in nontransformed cells. To determine whether induction of aneuploidy, tetraploidy, and G2 arrest by E. faecalis in HCT116 and RKO cells extended to nontransformed cells, we exposed YAMC cells to E. faecalis. A similar set of responses was found with increased aneuploidy and tetraploidy, γH2AX formation, decreased γH2AX foci after treatment with MnSOD and catalase, and G2 arrest (Supplementary Fig. S5). Caffeine decreased γH2AX foci (data not shown) and partially rescued YAMC cells from E. faecalis–induced G2 arrest (Supplementary Fig. S5). These findings confirmed that the DNA-damaging and cyclomodulatory effects of E. faecalis were not unique to transformed cells.

E. faecalis–infected macrophages induce aneuploidy through a bystander effect. Colonic mucus limits direct contact between intestinal commensals and epithelial cells. Thus, luminal bacteria would likely need to exert promutagenic effects through indirect mechanisms. For E. faecalis, this could occur via translocation of the intact colonic epithelium (39). We have shown in mice that this process can lead to acute changes in colonic mucosal gene expression and nuclear factor-κB activation in tissue macrophages (27). This finding is consistent with other work by us, showing activated macrophages produce clastogens that damage epithelial cells via a mechanism analogous to the radiation-induced bystander effect (9, 13). To investigate this hypothesis, we exposed HCT116 cells to murine macrophages infected with E. faecalis in a dual-chamber tissue culture system (9). We found the percentage of aneuploid and tetraploid cells increased significantly at an MOI of 1,000 compared with uninfected controls and at lower MOIs (Fig. 4A).

Figure 4.

E. faecalis–infected macrophages induce aneuploidy and tetraploidy in colonic epithelial cells through a bystander effect. A, macrophages infected with E. faecalis at an MOI of 1,000 increase the percentage of aneuploid and tetraploid HCT116 cells. ###, P < 0.001 compared with coculture with uninfected macrophages. NS, not significant. B, depletion of GSH in HCT116 cells by 50 μmol/L BSO increases the percentage of aneuploid and tetraploid cells when cocultured with E. faecalis–infected macrophages at MOIs from 1 to 1,000. #, P < 0.05; ###, P < 0.001, compared with exposure to uninfected macrophages (MOI = 0); ***, P < 0.001, compared with without macrophages.

Figure 4.

E. faecalis–infected macrophages induce aneuploidy and tetraploidy in colonic epithelial cells through a bystander effect. A, macrophages infected with E. faecalis at an MOI of 1,000 increase the percentage of aneuploid and tetraploid HCT116 cells. ###, P < 0.001 compared with coculture with uninfected macrophages. NS, not significant. B, depletion of GSH in HCT116 cells by 50 μmol/L BSO increases the percentage of aneuploid and tetraploid cells when cocultured with E. faecalis–infected macrophages at MOIs from 1 to 1,000. #, P < 0.05; ###, P < 0.001, compared with exposure to uninfected macrophages (MOI = 0); ***, P < 0.001, compared with without macrophages.

Close modal

Eukaryotic cells have numerous defenses against oxidative stress and maintenance of genomic integrity. Glutathione (GSH) is a cofactor for enzymes that protect cells against free radical toxicity and xenobiotics. To determine whether GSH-dependent antioxidant defenses protected against the bystander effect, we depleted HCT116 cells of intracellular GSH using buthionine sulfoximine (BSO), a specific inhibitor of GSH synthesis. Depleted cells were exposed to E. faecalis–infected macrophages in the dual-chamber culture system. As the MOI increased from 1 to 100, the proportion of aneuploid and tetraploid cells increased compared with uninfected macrophages (Fig. 4B). Cells treated with BSO alone showed no increase in the proportion of aneuploid and tetraploid cells. A slight increase was seen, however, in GSH-depleted cells cocultured with uninfected macrophages. This may represent basal production of clastogens by these cells.4

4

M.M. Huycke, unpublished data.

Thus, GSH-dependent cellular defenses protect against clastogen-mediated aneugenesis and support the notion that E. faecalis-infected macrophages induce CIN through a bystander effect.

Infected macrophages cause DSBs through a bystander effect. To assess the response of colonic epithelial cells to E. faecalis–infected macrophages, we investigated DNA damage repair pathways using the dual-chamber coculture system. Cells were synchronized in G1 by double thymidine block and exposed to E. faecalis–infected macrophages at an MOI of 1,000. HCT116 cells developed a marked G2-M arrest (Fig. 5A,, left). The percentage of cells in G2-M was significantly increased at 48 hours (8.0 ± 0.1) and 72 hours (9.6 ± 0.5) compared with cells exposed to uninfected macrophages. In comparison, the percentage of RKO cells in G2-M did not change significantly (<5%; Fig. 5A , right).

Figure 5.

E. faecalis–infected macrophages induce DSBs in colonic epithelial cells through a bystander effect. A, exposure of synchronized HCT116 cells to E. faecalis–infected macrophages at an MOI of 1,000 induces a G2-M arrest after 48 h compared with exposure to uninfected macrophages (left). Compared with coculture with uninfected macrophages, no significant change in cell cycle is seen for RKO cells exposed to E. faecalis–infected macrophages at an MOI of 1,000 (right). B, coculture with uninfected or E. faecalis–infected macrophages activates ATM/ATR DNA damage responses in HCT116 cells. Cyclin B1 is increased at 5, 10, and 24 h postexposure to infected macrophages compared with uninfected controls. C, laser scanning confocal microscopy shows γH2AX foci (green) in nuclei (blue) of HCT116 cells treated with supernatants from overnight cultures of E. faecalis–infected macrophages at an MOI of 1,000 (bottom). Similar foci were rare for cells treated with supernatants from uninfected macrophages (top). D, FACS analysis shows increased percentage of HCT116 cells (left) and RKO cells (right) with γH2AX foci after exposure to E. faecalis–infected macrophages at an MOI of 1,000 at 48 h compared with uninfected control (4.3% versus 15.5% and 8.7% versus 16.1%, respectively). Gray trace, untreated control; red trace, coculture with uninfected control; green trace, coculture with E. faecalis–infected macrophages.

Figure 5.

E. faecalis–infected macrophages induce DSBs in colonic epithelial cells through a bystander effect. A, exposure of synchronized HCT116 cells to E. faecalis–infected macrophages at an MOI of 1,000 induces a G2-M arrest after 48 h compared with exposure to uninfected macrophages (left). Compared with coculture with uninfected macrophages, no significant change in cell cycle is seen for RKO cells exposed to E. faecalis–infected macrophages at an MOI of 1,000 (right). B, coculture with uninfected or E. faecalis–infected macrophages activates ATM/ATR DNA damage responses in HCT116 cells. Cyclin B1 is increased at 5, 10, and 24 h postexposure to infected macrophages compared with uninfected controls. C, laser scanning confocal microscopy shows γH2AX foci (green) in nuclei (blue) of HCT116 cells treated with supernatants from overnight cultures of E. faecalis–infected macrophages at an MOI of 1,000 (bottom). Similar foci were rare for cells treated with supernatants from uninfected macrophages (top). D, FACS analysis shows increased percentage of HCT116 cells (left) and RKO cells (right) with γH2AX foci after exposure to E. faecalis–infected macrophages at an MOI of 1,000 at 48 h compared with uninfected control (4.3% versus 15.5% and 8.7% versus 16.1%, respectively). Gray trace, untreated control; red trace, coculture with uninfected control; green trace, coculture with E. faecalis–infected macrophages.

Close modal

Activation of the G2 checkpoint by DSBs can occur through protein kinase damage response pathways. p-ATM/p-Chk2 and p-ATR/p-Chk1 signaling pathways were activated in both cell lines exposed to uninfected macrophages (Fig. 5B and Supplementary Fig. S6). Activation was moderately stronger, however, after exposure to E. faecalis–infected macrophages. As previously shown for cells directly exposed to E. faecalis (Fig. 2B and C), phosphorylation of Cdk1 at Tyr15 was evident at baseline and did not increase after exposure to E. faecalis–infected macrophages. A G2-M arrest, however, was only noted for HCT116 cells (Fig. 5A), suggesting differences in sensitivity to DNA damage or cell cycle checkpoints between these cell types.

We next identified γH2AX in cells exposed to supernatants from E. faecalis–infected macrophages (Fig. 5C,, bottom, and Supplementary Fig. S7). By laser scanning confocal microscopy γH2AX foci appeared more frequently in cells exposed to supernatants from infected macrophages than uninfected macrophages. Similarly, the proportion of γH2AX-positive cells at 5 hours (data not shown) and 48 hours (Fig. 5D) after exposure to infected macrophages was significantly greater for cells exposed to uninfected macrophages. The proportion of γH2AX-positive RKO cells (Fig. 5D,, right) was less than that of HCT116 cells (Fig. 5D , left), suggesting insufficient DNA damage as a potential explanation for the lack of arrest at G2. These data showed that the bystander effect can activate ATM and ATR, generate γH2AX foci, and, for HCT116 cells, induce a G2-M arrest.

E. faecalis induces γH2AX foci in vivo. To determine whether E. faecalis could generate γH2AX foci in vivo, we developed an intestinal ligation model that exposed the colonic mucosa to an inoculum of enterococci. In this system, wild-type mice were exposed to E. faecalis for 6 hours, PBS as sham, or irradiated with 13 Gy. There was no histologic evidence for inflammation or mucosal abnormality in control or E. faecalis–exposed mice. However, immunohistochemistry showed marked positivity for γH2AX in colonic epithelial and stromal cell nuclei for mice treated with E. faecalis or irradiated (Fig. 6A,, C, and D). Comparison of E. faecalis–exposed colons to shams showed significant increases in γH2AX in the nuclei of epithelial and stromal cells (Fig. 6B–D). The average (±SD) number of positive nuclei per cross-section for surface epithelial, crypt epithelial, and stromal cells for E. faecalis–exposed compared with sham mice was 2.3 ± 1.7, 12.0 ± 7.2, and 3.1 ± 2.2 versus 0.3 ± 0.5, 0.7 ± 1.2, and 0.0 ± 0.0, respectively (P < 0.001 for all comparisons). These results show that E. faecalis can generate rapid induction of γH2AX and, presumably, DSBs in colonic epithelial and stromal cells—findings not dissimilar to radiation-induced intestinal injury.

Figure 6.

E. faecalis induces DSBs in vivo. A, colons of wild-type mice show nuclei with numerous epithelial and stromal cells positive for γH2AX by immunohistochemistry after 13-Gy irradiation (arrows, 60×). B, colonic mucosa from the ligation model for PBS controls show few nuclei staining for γH2AX (60×). C, colonic mucosa from mice exposed to E. faecalis show numerous foci of nuclear staining for γH2AX in epithelial and stromal cells (arrows, 60×). D, same as C showing nuclear γH2AX staining in stromal cell adjacent to colonic crypt (arrow, 60×).

Figure 6.

E. faecalis induces DSBs in vivo. A, colons of wild-type mice show nuclei with numerous epithelial and stromal cells positive for γH2AX by immunohistochemistry after 13-Gy irradiation (arrows, 60×). B, colonic mucosa from the ligation model for PBS controls show few nuclei staining for γH2AX (60×). C, colonic mucosa from mice exposed to E. faecalis show numerous foci of nuclear staining for γH2AX in epithelial and stromal cells (arrows, 60×). D, same as C showing nuclear γH2AX staining in stromal cell adjacent to colonic crypt (arrow, 60×).

Close modal

In this study, we found that the human intestinal commensal E. faecalis induced γH2AX foci in human colonic epithelial cells, activated DNA damage pathways, produced G2 arrest, and promoted missegregation of chromosomes, leading to aneuploidy and tetraploidy. The production of ROS by E. faecalis was associated with anaphase bridging, lagging chromosomes, and multipolar mitoses. MnSOD, catalase, and tocopherols attenuated aneugenic and cyclomodulatory effects, linking the redox-active phenotype of this commensal to these potentially transforming events. E. faecalis infection of macrophages produced a bystander effect that resulted in γH2AX deposition, cell cycle arrest, and CIN. Finally, the colonic mucosa of mice exposed to E. faecalis showed increased γH2AX formation similar to that produced by irradiation. γH2AX formation is a marker of DSBs, and these findings support our previous work that E. faecalis can generate DNA damage in vivo (6). In general, DSBs are difficult to repair and can generate dicentric chromosomes through nonhomologous end-joining repair (28). Dicentric chromosomes can attach to both spindle poles and lead to anaphase bridging. If bridges fragment, chromosomes can enter breakage-fusion-bridge cycles that produce rearrangements and aneuploidy (28). These are hallmark features of CIN.

We selected E. faecalis–exposed colonic epithelial cells with >5N DNA content to determine if aneuploid cells could persist and found that few survived clonal expansion. Of those that propogated, all reverted to near-diploid, except for one clone that remained tetraploid. Indeed, the production of ROS by E. faecalis was associated with a significantly increased number of tetraploid cells. Polyploid cells with more than two sets of homologous chromosomes can, in general, arise after cellular stress through a failure of cytokinesis by cell fusion or from mitotic slippage (40). Tetraploid cells, in particular, are considered potential intermediates in the cellular progression from diploid to aneuploid. For example, supernumerary centrosomes in tetraploid cells can form multipolar mitoses and cause haphazard chromosome segregation. The tumorigenic potential of tetraploid cells was recently confirmed using a p53-deficient primary mouse mammary epithelial cell that, when exposed to cytochalasin, was rendered tetraploid and found to be malignant in nude mice (41).

E. faecalis–induced G2 arrest was partially ameliorated by MnSOD and indicated that superoxide can act as a cyclomodulin. Cyclomodulins are a family of secreted bacterial toxins that interfere with the eukaryotic cell cycle (42). One example, produced by extraintestinal pathogenic and commensal strains of Escherichia coli, is the cytolethal distending toxin that induces DSBs, cell distension, G2 checkpoint activation, and, ultimately, cell death (25). Commensal bacteria with cyclomodulatory effects may initiate DSBs that lead to anaphase bridging, chromosomal fragmentation, and chromosomal rearrangements. G2 arrest should prevent cell cycle progression until DNA damage is repaired. However, damage below threshold values of 10 to 20 DSBs needed for the activation of mammalian checkpoints, would lead to propagation of chromosome aberrations to daughter cells (43). Evidence for this has been noted in sporadic CRC with mutations arising via low-level, telomere-mediated anaphase bridging (23).

G2 arrest is typically induced by decreased dephosphorylation of Cdk1 at Tyr15/The14 (44). However, by Western blot, we found no evidence for Cdk1 inactivation in HCT116 or RKO cells exposed to E. faecalis. HCT116 and RKO cells are mismatch repair–deficient, and this trait can negatively affect G2 checkpoint activation (45). Cell cycle arrest during G2 was confirmed by the persistence of cyclin B1 and the lack of accumulation of cells in mitosis. However, classic activation of the G2 checkpoint through Cdk1 was not found, and the G2 arrest that was induced by E. faecalis may have occurred through potential alternate pathways, including p53/p21, GADD45, and 14-3-3σ, among others (45, 46). This is a subject of ongoing investigation.

We previously found that E. faecalis promoted CIN in colonic epithelial cells through a macrophage-induced bystander effect (9). As with radiation, superoxide is a trigger for this effect (47). COX-2 contributes to both the radiation-induced and macrophage-induced bystander effects (9, 15, 16), although mechanisms are unclear. The results reported here show that the macrophage-induced bystander effect can rapidly induce aneuploidy and tetraploidy in colonic epithelial cells. The bystander effect also generated γH2AX foci in target cells, a finding consistent with previous studies showing chromosomal DSBs due to clastogenesis (47). Depleting intracellular GSH with BSO amplified these effects, suggesting that antioxidant or xenobiotic defenses can protect against clastogenesis and aneugenesis. Several potential mediators for the bystander effect include tumor necrosis factor-α and breakdown products of polyunsaturated fatty acids, such as 4-hydroxy-2-nonenal (47, 48), although no single agent has been definitively shown to be causative.

In conclusion, this study shows that E. faecalis can induce γH2AX foci, G2 arrest, anaphase bridging, and aneuploidy and tetraploidy in human colonic epithelial cells through direct contact or after exposure to E. faecalis–infected macrophages. The latter is analogous to the radiation-induced bystander effect (13). Blocking DNA repair with caffeine or depleting GSH-dependent defenses with BSO significantly increased the frequency of aneuploidy and tetraploidy. Finally, E. faecalis–induced γH2AX formation in murine colons suggested that this commensal can generate DSBs in vivo. These findings provide insights into mechanisms by which commensal bacteria may contribute to the endogenous origin of CIN and cellular transformation.

No potential conflicts of interest were disclosed.

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

Grant support: Department of Veterans Affairs, NIH CA127893 (M.M. Huycke), and Frances Duffy Endowment.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank the assistance of Danny Moore, Jim Henthorn (Flow Cytometry Laboratory at the University of Oklahoma Health Sciences Center), Shibo Li (Children's Hospital of Oklahoma), and Susannah Rankin (Oklahoma Medical Research Foundation).

1
Hope ME, Hold GL, Kain R, El-Omar EM. Sporadic colorectal cancer-role of the commensal microbiota.
FEMS Microbiol Lett
2005
;
244
:
1
–7.
2
Moore WEC, Moore LH. Intestinal floras of populations that have a high risk of colon cancer.
Appl Environ Microbiol
1995
;
61
:
3202
–7.
3
IARC. Chronic infections. In: Stewart BW, Kleihues P, editors. World Cancer Report. Lyon (France): IARC Press; 2003. p. 56–61.
4
Eckburg PB, Bik EM, Bernstein CN, et al. Diversity of the human intestinal microbial flora.
Science
2005
;
308
:
1635
–8.
5
Huycke MM, Moore D, Shepard L, et al. Extracellular superoxide production by Enterococcus faecalis requires demethylmenaquinone and is attenuated by functional terminal quinol oxidases.
Mol Microbiol
2001
;
42
:
729
–40.
6
Huycke MM, Abrams V, Moore DR. Enterococcus faecalis produces extracellular superoxide and hydrogen peroxide that damages colonic epithelial cell DNA.
Carcinogenesis
2002
;
23
:
529
–36.
7
Kim SC, Tonkonogy SL, Albright CA, et al. Variable phenotypes of enterocolitis in IL-10 deficient mice monoassociated with two different commensal bacteria.
Gastroenterology
2005
;
128
:
891
–906.
8
Balish E, Warner T. Enterococcus faecalis induces inflammatory bowel disease in interleukin-10 knockout mice.
Am J Pathol
2002
;
160
:
2253
–7.
9
Wang X, Huycke MM. Extracellular superoxide production by Enterococcus faecalis promotes chromosomal instability in mammalian cells.
Gastroenterology
2007
;
132
:
551
–61.
10
Jallepalli PV, Lengauer C. Chromosome segregation and cancer: cutting through the mystery.
Nat Rev Cancer
2001
;
1
:
109
–17.
11
Shih I-M, Zhou W, Goodman SN, Lengauer C, Kinzler KW, Vogelstein B. Evidence that genetic instability occurs at an early stage of colorectal tumorigenesis.
Cancer Res
2001
;
61
:
818
–22.
12
Beckman RA, Loeb LA. Genetic instability in cancer: theory and experiment.
Semin Cancer Biol
2005
;
15
:
423
–35.
13
Lorimore SA, Wright EG. Radiation-induced genomic instability and bystander effects: related inflammatory-type responses to radiation-induced stress and injury? A review.
Int J Radiat Biol
2003
;
79
:
15
–25.
14
Watson GE, Lorimore SA, Macdonald DA, Wright EG. Chromosomal instability in unirradiated cells induced in vivo by a bystander effect of ionizing radiation.
Cancer Res
2000
;
60
:
5608
–11.
15
Zhou H, Ivanov VN, Gillespie J, et al. Mechanism of radiation-induced bystander effect: role of the cyclooxygenase-2 signaling pathway.
Proc Natl Acad Sci U S A
2005
;
102
:
14641
–6.
16
Lee SH, Williams MV, Dubois RN, Blair IA. Cyclooxygenase-2-mediated DNA damage.
J Biol Chem
2005
;
280
:
28337
–46.
17
Tlsty TD, Coussens LM. Tumor stroma and regulation of cancer development.
Annu Rev Pathol
2006
;
1
:
119
–50.
18
Lengauer C, Kinzler KW, Vogelstein B. Genetic instability in colorectal cancers.
Nature
1997
;
386
:
623
–7.
19
Roschke AV, Stover K, Tonon G, Schaffer AA, Kirsch IR. Stable karyotypes in epithelial cancer cell lines despite high rates of ongoing structural and numerical chromosomal instability.
Neoplasia
2002
;
4
:
19
–31.
20
Whitehead RH, VanEeden PE, Noble MD, Ataliotis P, Jat PS. Establishment of conditionally immortalized epithelial cell lines from both colon and small intestine of adult H-2Kb-tsA58 transgenic mice.
Proc Natl Acad Sci U S A
1993
;
90
:
587
–91.
21
Huycke MM, Moore DR. In vivo production of hydroxyl radical by Enterococcus faecalis colonizing the intestinal tract using aromatic hydroxylation.
Free Radic Biol Med
2002
;
33
:
818
–26.
22
Noble CJ. Carriage of group D streptococci in the human bowel.
J Clin Pathol
1978
;
31
:
1182
–6.
23
Stewenius Y, Gorunova L, Jonson T, et al. Structural and numerical chromosome changes in colon cancer develop through telomere-mediated anaphase bridges, not through mitotic multipolarity.
Proc Natl Acad Sci U S A
2005
;
102
:
5541
–6.
24
Muehlbauer PA, Schuler MJ. Detection of numerical chromosomal aberrations by flow cytometry: a novel process for identifying aneugenic agents.
Mutat Res
2005
;
585
:
156
–69.
25
Nougayrede JP, Homburg S, Taieb F, et al. Escherichia coli induces DNA double-strand breaks in eukaryotic cells.
Science
2006
;
313
:
848
–51.
26
Gentry-Weeks CR, Karkhoff-Schweizer R, Pikis A, Estay M, Keith JM. Survival of Enterococcus faecalis in mouse peritoneal macrophages.
Infect Immun
1999
;
67
:
2160
–5.
27
Allen TD, Moore DR, Wang X, et al. Dichotomous metabolism of Enterococcus faecalis induced by hematin starvation modulates colonic gene expression.
J Med Microbiol
2008
;
57
:
1193
–204.
28
Acilan C, Potter DM, Saunders WS. DNA repair pathways involved in anaphase bridge formation.
Genes Chromosomes Cancer
2007
;
46
:
522
–31.
29
Hensley K, Benaksas EJ, Bolli R, et al. New perspectives on vitamin E: γ-tocopherol and carboxyelthylhydroxychroman metabolites in biology and medicine.
Free Radic Biol Med
2004
;
36
:
1
–15.
30
Bayani J, Selvarajah S, Maire G, et al. Genomic mechanisms and measurement of structural and numerical instability in cancer cells.
Semin Cancer Biol
2007
;
17
:
5
–18.
31
Su TT. Cellular responses to DNA damage: one signal, multiple choices.
Annu Rev Genet
2006
;
40
:
187
–208.
32
Matsuoka S, Huang M, Elledge SJ. Linkage of ATM to cell cycle regulation by the Chk2 protein kinase.
Science
1998
;
282
:
1893
–7.
33
Chaturvedi P, Eng WK, Zhu Y, et al. Mammalian Chk2 is a downstream effector of the ATM-dependent DNA damage checkpoint pathway.
Oncogene
1999
;
18
:
4047
–54.
34
Yarden RI, Pardo-Reoyo S, Sgagias M, Cowan KH, Brody LC. BRCA1 regulates the G2/M checkpoint by activating Chk1 kinase upon DNA damage.
Nat Genet
2002
;
30
:
285
–9.
35
Aoki K, Aoki M, Sugai M, et al. Chromosomal instability by β-catenin/TCF transcription in APC or β-catenin mutant cells.
Oncogene
2007
;
26
:
3511
–20.
36
Bartkova J, Hořejsi Z, Koed K, et al. DNA damage response as a candidate anti-cancer barrier in early human tumorigenesis.
Nature
2005
;
434
:
864
–70.
37
Böhm L, Roos WP, Serafin AM. Inhibition of DNA repair by pentoxifylline and related methylxanthine derivatives.
Toxicology
2003
;
193
:
153
–60.
38
Pero RW, Anderson MW, Doyle GA, et al. Oxidative stress induces DNA damage and inhibits the repair of DNA lesions induced by N-acetoxy-2-acetylaminofluorene in human peripheral mononuclear leukocytes.
Cancer Res
1990
;
50
:
4619
–25.
39
Wells CL, Erlandsen SL. Localization of translocating Escherichia coli, Proteus mirabilis, and Enterococcus faecalis within cecal and colonic tissues of monoassociated mice.
Infect Immun
1991
;
59
:
4693
–7.
40
Ganem NJ, Storchova Z, Pellman D. Tetraploidy, aneuploidy and cancer.
Curr Opin Genet Dev
2007
;
17
:
157
–62.
41
Fujiwara T, Bandi M, Nitta M, Ivanova EV, Bronson RT, Pellman D. Cytokinesis failure generating tetraploids promotes tumorigenesis in p53-null cells.
Nature
2005
;
437
:
1043
–7.
42
Oswald E, Nougayrede JP, Taieb F, Sugai M. Bacterial toxins that modulate host cell-cycle progression.
Curr Opin Microbiol
2005
;
8
:
83
–91.
43
Löbrich M, Jeggo PA. The impact of a negligent G2/M checkpoint on genomic instability and cancer induction.
Nat Rev Cancer
2007
;
7
:
861
–9.
44
Poon RY, Jiang W, Toyoshima H, Hunter T. Cyclin-dependent kinases are inactivated by a combination of p21 and Thr-14/Tyr-15 phosphorylation after UV-induced DNA damage.
J Biol Chem
1996
;
271
:
13283
–91.
45
Yan T, Schupp JE, Hwang HS, et al. Loss of DNA mismatch repair imparts defective cdc2 signaling and G2 arrest responses without altering survival after ionizing radiation.
Cancer Res
2001
;
61
:
8290
–7.
46
Taylor WR, Stark GR. Regulation of the G2/M transition by p53.
Oncogene
2001
;
20
:
1803
–15.
47
Emerit I. Reactive oxygen species, chromosome mutation, and cancer: possible role of clastogenic factors in carcinogenesis.
Free Radic Biol Med
1994
;
16
:
99
–109.
48
Yan B, Wang H, Rabbani ZN, et al. Tumor necrosis factor-α is a potent endogenous mutagen that promotes cellular transformation.
Cancer Res
2006
;
66
:
11565
–70.

Supplementary data