Vancomycin-sensitive bacteria trigger development of colitis-associated colon cancer by attracting neutrophils

Inflammatory bowel disease confers an increased risk of developing colitis-associated colon cancer (CAC). During the active colitis or developing tumor stage, commensal bacteria show dynamic translocation. However, whether alteration of the bacterial composition in the gut causes CAC is still unclear. To clarify the effect of commensal bacteria on CAC development, we employed an azoxymethane (AOM) and dextran sodium sulfate (DSS)-induced murine CAC model treated with or without antibiotics. In addition, we analyzed the effects of antibiotics on infiltration of myeloid cells, colonic inflammatory responses, and colorectal cancer formation. We found that vancomycin treatment dramatically suppressed tumor development. In addition, AOM/DSS treatment greatly induced the infiltration of Gr-1high/CD11bhigh neutrophils to the colon, which led to the production of tumor necrosis factor α and inducible nitric oxide synthase. Vancomycin treatment suppressed the infiltration of neutrophils induced by AOM/DSS. Moreover, vancomycin treatment greatly reduced the colon injury and DNA damage caused by AOM/DSS-induced NO radicals. Our results indicate that vancomycin-sensitive bacteria induced colon inflammation and DNA damage by attracting neutrophils into damaged colon tissue, thus promoting tumor formation.


Results
Vancomycin reduced the risk of colitis-associated carcinogenesis. To assess the effect of antibiotics on colitis and colon carcinogenesis, we employed a CAC model. The carcinogen AOM was injected in C57BL/6 mice followed by three rounds of 2% dextran sodium sulfate (DSS) in drinking water. Throughout the AOM/ DSS treatment, mice were orally administrated a combination of vancomycin and neomycin as broad-spectrum antibiotics or either vancomycin or neomycin alone. Tumor formation was analyzed at day 67 of AOM/DSS treatment (Fig. 1A). In the absence of antibiotic treatment, AOM/DSS-treated mice exhibited a high tumor burden in the colon-rectum regions at day 67 (Fig. 1B,C). While neomycin treatment alone did not affect tumor formation, vancomycin with or without neomycin markedly reduced AOM/DSS-induced CAC development (Fig. 1B,C). To analyze the pathology of AOM/DSS-induced tumors, we performed hematoxylin and eosin (H&E) staining of colon tissue at day 67 of AOM/DSS treatment. Histological analysis showed that vancomycin/neomycin or vancomycin alone greatly suppressed the development of adenocarcinoma induced by AOM/DSS treatment (Fig. 1D). Immunohistochemical evaluation confirmed that the activation of β -catenin and c-myc in the AOM/DSStreated colon epithelia was reduced by vancomycin/neomycin or vancomycin alone ( Supplementary Fig. S1A). Vancomycin treatment also decreased the epithelial proliferation in AOM/DSS-treated mice as shown by immunohistochemical analysis of cyclin D1 and Ki-67 ( Supplementary Fig. S1A,B). Moreover, we found that the expression of the phosphorylated histone p-H2AX, a marker for double-stranded DNA breaks, was increased in AOM/DSS-treated control colons and neomycin-treated colons, and its expression was greatly reduced by vancomycin/neomycin or vancomycin treatment ( Fig. 2A,B). The expression of inducible nitric oxide synthase (iNOS), apoptosis-related protein Bax, Bcl2, and phosphorylated p53 was induced by AOM/DSS treatment, and their expression was suppressed by vancomycin but not neomycin (Fig. 2C). These results indicate that vancomycin inhibits DNA damage, apoptosis, and proliferation of epithelial cells, and development of CAC.
Vancomycin reduced the infiltration of neutrophils to colon tumors. To further delineate the effect of vancomycin treatment on CAC development, we evaluated the migration ability of myeloid cells such as neutrophils (Gr-1 high /CD11b high ), dendritic cells (CD11c + /CD11b + ), and macrophages (F4/80 + /CD11b + ) to colon cancer induced by AOM/DSS treatment with or without vancomycin. At day 67 after AOM/DSS treatment, we observed that a large number of neutrophils were recruited to the AOM/DSS-treated colon. Vancomycin treatment greatly reduced the infiltration of neutrophils, although we observed a small remaining number of macrophages and dendritic cells (Fig. 3A,B). In bone marrow, although CD11c + /CD11b + dendritic cells and F4/80 + /CD11b + macrophages were not changed by AOM/DSS treatment, Gr-1 high /CD11b high neutrophils were enhanced at day 67 of AOM/DSS treatment, and vancomycin treatment reduced the expansion of these neutrophils ( Supplementary Fig. S2). To understand whether the number of neutrophils was increased by tumors or by colitis, we analyzed the number of neutrophils (Gr-1 high /CD11b high ) during the course of AOM/DSS-induced carcinogenesis in colon, bone marrow, and spleen by flow cytometric analysis. From the onset of treatment, AOM/DSS markedly increased the number of neutrophils not only in colon but also in bone marrow and spleen ( Supplementary Fig. S3). These results suggest that AOM/DSS treatment induces long-term infiltration of neutrophils into the colon, which might prolong inflammation and promote colon carcinogenesis. Furthermore, bone marrow and spleen appear to continuously prepare neutrophil precursors and contribute to colon inflammation. These data provide the evidence that vancomycin reduces the neutrophil migration from bone marrow and spleen to the inflamed colon in the development of CAC.
Vancomycin treatment attenuated colon inflammation. We next measured the expression of mRNA for chemokines and pro-inflammatory mediators in the colon using real-time PCR analysis. AOM/DSS-treated mice showed elevated expression of the chemokines CXCL1, CXCL2, and CCL2, and vancomycin treatment inhibited the expression of these chemokines (Fig. 3C). Moreover, we observed higher mRNA expression of the pro-inflammatory mediators tumor necrosis factor α (TNFα ), interleukin-6 (IL-6), and iNOS in AOM/ DSS-treated colons. The mRNA levels of TNFα and iNOS were significantly downregulated by vancomycin treatment (Fig. 3C). The expression of IL-6 appeared to be downregulated by vancomycin treatment but was not significant (Fig. 3C). The expression of regenerating islet-derived protein 3 gamma (RegIIIγ ), which is related to tissue damage and repair, was also strongly induced in AOM/DSS-treated colons, and vancomycin treatment significantly reduced its expression (Fig. 3C). Taken together, these data suggest that vancomycin-sensitive bacteria caused recruitment of neutrophils, which promoted TNFα and iNOS expression, followed by epithelial tissue damage, DNA damage, and tumor growth.
Vancomycin could decrease the severity of colitis. Because we observed that neutrophil infiltration and inflammation were induced in the colon at day 67 after AOM/DSS treatment, which was reduced by vancomycin, next we examined the contribution of vancomycin-sensitive bacteria in the colon early in the time course of AOM/DSS treatment (Fig. 4A). At day 10, a shorter colon length was found in AOM/DSS-treated mice (Fig. 4B). Administration of vancomycin/neomycin or vancomycin alone could inhibit the colon shortage induced by AOM/DSS, whereas treatment with neomycin alone did not (Fig. 4B). Histological analysis also revealed that the combination of vancomycin/neomycin or vancomycin alone reduced colon inflammation, such as ulceration and infiltration, but that neomycin did not affect the severity of colitis (Fig. 4C,D). These results further support the idea that vancomycin-sensitive bacteria play a critical role in the severity of colitis.
Scientific RepoRts | 6:23920 | DOI: 10.1038/srep23920 We also examined the expression of pro-inflammatory mediators at day 10 (inflammatory phase) of AOM/ DSS treatment. Vancomycin/neomycin or vancomycin alone greatly downregulated iNOS and TNFα mRNA expression, but not that of IL-6 and IL-1β at day 10 ( Fig. 5A). Because iNOS catalyzes the production of toxic reactive nitrogen species and elevates DNA damage and apoptosis, we examined the production of reactive oxygen species (ROS) in the AOM/DSS-treated colons. ROS production was reduced by vancomycin/neomycin treatment (Fig. 5B). Moreover, the DNA damage marker p-H2AX was highly induced at day 10 of AOM/DSS treatment and was decreased by treatment with vancomycin/neomycin or vancomycin alone (Fig. 5C,D). These results suggest that vancomycin-sensitive bacteria induce iNOS and TNFα production in damaged colon tissue, which leads to DNA damage and apoptosis in colon epithelial cells.

Vancomycin inhibited the recruitment of neutrophils into AOM/DSS-treated colon tissue. As
we had observed neutrophil infiltration to the colon at day 67 of AOM/DSS treatment, we evaluated neutrophil infiltration to AOM/DSS-treated colons at the earlier time point (day 10). AOM/DSS-treated mice showed an elevated expression of the chemokines CXCL1 and CXCL2, which induce neutrophil recruitment, whereas vancomycin treatment inhibited the expression of these chemokines (Fig. 6A). Indeed, AOM/DSS treatment strongly induced the infiltration of neutrophils at day 10 of AOM/DSS treatment (Fig. 6B,C). Vancomycin administration to the AOM/DSS-treated mice greatly reduced the infiltration of neutrophils to the colon at day 10, but neomycin treatment did not (Fig. 6B,C). AOM/DSS-induced tissue injury and colon infiltration with neutrophils were sustained until day 24 ( Supplementary Fig. S4A-C). In bone marrow and spleen, Gr-1 high /CD11b high neutrophils at day 24 were enhanced by AOM/DSS treatment ( Supplementary Fig. S5A,B). We confirmed that the Gr-1 high /CD11b high cells that were induced by AOM/DSS treatment were neutrophils with ring-shaped nuclei ( Supplementary Fig. S6A). They produced ROS, iNOS, and several pro-inflammatory cytokines in colitis ( Supplementary Fig. S6B,C).
Because of the observed effects of the antibiotic vancomycin on AOM/DSS-treated colon tissues, we examined the bacterial composition of the colon at day 24 of AOM/DSS treatment using real-time PCR to analyze bacterial 16S ribosomal DNA. AOM/DSS-treated mice exhibited a marked increase in the amount of Gram-positive bacteria, including Clostridium leptum, C. coccoides, and segmented filamentous bacteria (SFB), as well as Gram-negative Enterobacteriaceae in their stool (Fig. 6D). We confirmed that vancomycin treatment dramatically reduced the amount of Gram-positive bacteria, including Faecalibacterium prausnitzii, C. leptum, C. coccoides and SFB. In addition, vancomycin treatment reduced the number of the anaerobic Gram-negative bacteria Bacteroides/Prevotella and mouse intestinal Bacteroides (MI Bacteroides), while neomycin treatment did not. On the other hand, the number of aerobic Gram-negative Enterobacteriaceae was significantly reduced by treatment with neomycin but not vancomycin (Fig. 6D). In addition, levels of Gram-positive Enterococcus and Lactobacillus spp. and anaerobic Gram-negative Fusobacterium spp. were slightly increased by AOM/DSS treatment, but they were not significantly reduced by vancomycin or neomycin (Fig. 6D). These results suggest that vancomycin-sensitive bacteria, such as C. leptum, C. coccoides, SFB, and Bacteroides, in the colon might trigger the migration of neutrophils to the inflamed colon from bone marrow and spleen. The neutrophil-produced iNOS then causes epithelial DNA damage, which in turn promotes colon carcinogenesis.

Aminoguanidine (AG) treatment inhibited the initiation of CAC.
To investigate the effect of iNOS on colitis and colitis-associated carcinogenesis, we treated mice with the iNOS inhibitor AG upon injection of AOM (12 mg/kg), followed by three rounds of DSS treatment (Fig. 7A). At day 10, AG-treated mice exhibited colon shortness similar to the AOM/DSS-treated control mice (Fig. 7B). Furthermore, H&E staining of the colon at day 10 revealed that AG-treated mice also showed severe colon tissue damage (Fig. 7C). By flow cytometric analysis, we also found that the infiltration of neutrophils in AOM/DSS-treated colons was similar regardless of AG treatment (Fig. 7D,E). However, at the end of AOM/DSS treatment (at day 67), AOM/DSS/AG-treated mice had fewer tumors compared with AOM/DSS-treated control mice (Fig. 7F-H). These results indicate that iNOS has no effect on the severity of colitis but promotes AOM/DSS-induced carcinogenesis.
Taken together, vancomycin treatment suppressed the recruitment of Gr-1 high /CD11b high neutrophils caused by AOM/DSS treatment, and that reduced the severity of colitis and prevented CAC development. Hence, we suggest that vancomycin-sensitive bacteria trigger colitis and colon cancer in the murine CAC model.

Discussion
Here we showed that vancomycin dramatically suppressed inflammation-mediated tumorigenesis in the AOM/DSS-induced murine CAC model. We confirmed that vancomycin dramatically reduced the number of Gram-positive bacteria, such as F. prausnitzii, C. leptum, C. coccoides and SFB, as well as the anaerobic Gram-negative bacteria Bacteroides/Prevotella and MI Bacteroides, while neomycin did not. Thus, these results suggest that vancomycin-sensitive bacteria trigger colon inflammation and initiation of CAC. Our results may contrast with previous reports showing that Gram-negative bacteria Enterobacteriaceae, including Escherichia coli, induced inflammatory bowel disease such as Crohn's disease [26][27][28][29][30][31] . In our report, vancomycin treatment did not reduce the number of Enterobacteriaceae but dramatically reduced colon inflammation and carcinogenesis; neomycin, which kills Enterobacteriaceae, did not affect the severity of colitis or tumor multiplicity.   Our results showed that commensal bacteria induced the production of the chemokines CXCL1 and CXCL2, which induced the infiltration of Gr-1 high /CD11b high neutrophils in the AOM/DSS-induced colitis. In addition, we showed that during the course of AOM/DSS treatment, the number of Gr-1 high /CD11b high neutrophils increased greatly in bone marrow and spleen and appear to migrate to the injured colon. High expression of TNFα , iNOS, and ROS indicated that these neutrophils were pro-inflammatory. Previously, we demonstrated that Gr-1 high / CD11b high cells in the colon, bone marrow, and spleen were induced in an DSS-induced murine colitis model 22,23 . We have shown that infusion of neutrophils early in DSS-induced colitis reduces inflammation-induced tissue damage 22 , and the absence of neutrophils early in DSS-induced colitis worsens tissue damage 23 . Our previous reports have shown that Gr-1 high /CD11b high cells engulf translocated bacteria early in DSS-induced colitis, which contributes to early recovery from colitis 22,23 . In contrast to this complete recovery from colitis from one round of treatment with DSS, here with repeated AOM/DSS treatment we observed damage to colon tissue by prolonged infiltration of Gr-1 high /CD11b high neutrophils against the invasion of bacteria. Others have also reported that the absence of neutrophils reduced tumor multiplicity in the AOM/DSS-induced murine CAC model 32,33 . In these reports, Ly6G + /CD11b + cells (Gr-1 high /CD11b high cells) were recognized as myeloid-derived suppressor cells (MDSCs). Generally MDSCs have an M2-like phenotype, producing IL-10 and arginase. However, our neutrophils are pro-inflammatory cells that produce ROS, iNOS, and several pro-inflammatory cytokines in colitis ( Supplementary Fig. S6B,C). In this study, we showed that vancomycin treatment inhibited infiltration of neutrophils to the inflamed colon and suppressed colitis and colon carcinogenesis. Our results suggest that vancomycin-sensitive bacteria attract pro-inflammatory Gr-1 high /CD11b high neutrophils, which causes colitis and colon cancer.
The results here show that vancomycin treatment reduced TNFα and iNOS production in the colon. Recent studies using mouse models of inflammation-associated cancer, including CAC, have shown that inflammation mainly acts as tumor promoter 24,34,35 . Inflammatory mediators such as TNFα promote intestinal epithelial proliferation during CAC induction 24 . Erdman et al. reported that the presence of Gr-1 + neutrophils and elevated nitric oxide (NO) and TNFα trigger colonic inflammation and carcinogenesis in Rag2-deficient mice 36 . In our experiments, we demonstrated that vancomycin-sensitive bacteria induced infiltration of neutrophils, which produced iNOS and TNFα , and promoted inflammation-mediated tumorigenesis in the AOM/DSS-induced murine CAC model. Because this murine model is known to be similar to human ulcerative colitis, our results are valuable for understanding the mechanism of initiation of human ulcerative colitis and colon cancer.
In the murine model of colorectal cancer induced by loss of function of adenomatous polyposis coli (APC), it was reported that inhibition of two major enzymes, cyclooxygenase-2 (Cox2) and iNOS, suppressed tumor formation [37][38][39] . Also, experimental colitis was shown to increase tumorigenesis in APC+/Min mice through an iNOS-dependent mechanism 40 . iNOS is a primary regulator of NO production in innate immune cells, and NO contributes to bactericidal activities. NO reacts with superoxide radicals, resulting in formation of peroxynitrite, which generates nitrate 41 . Peroxynitrite not only kills bacteria, but also induces direct DNA damage in host intestinal epithelial cells, and enhances tumor initiation 39 . In our results, vancomycin treatment strongly reduced iNOS production and inhibited recruitment of iNOS-producing neutrophils induced by AOM/DSS treatment. Furthermore, colitis-induced colon tissue damage and DNA damage were dramatically suppressed by vancomycin treatment. We also confirmed the effect of iNOS in colitis and CAC development using the iNOS inhibitor AG. Previously, whether iNOS improves or exacerbates colitis was unclear 42 . Here, we clearly showed that iNOS did not affect the severity of colitis but played a role in initiation of colon cancer by contributing to DNA damage.
In conclusion, our results suggest that vancomycin-sensitive bacteria induced colon inflammation and DNA damage by attracting neutrophils to the damaged colon tissue, and finally, promoted tumor formation.

Materials and Methods
Mice. All experiments used C57BL/6 NCrSlc female mice, 6-8 weeks old, which were purchased from SLC (Japan SLC Inc., Shizuoka, Japan) and maintained at the Animal Research Facility at Nagoya University Graduate School of Medicine under specific pathogen-free conditions. This work was approved by the ethical committee of Nagoya University. All experiments were performed in accordance with the approved guidelines.
Induction of colorectal cancer. Mice were intraperitoneally injected with 12 mg/kg body weight of AOM (Sigma-Aldrich, St Louis, MO, USA) dissolved in phosphate-buffered saline (PBS). Five days after AOM administration, 2% DSS (MP Biomedicals, Santa Ana, CA, USA, molecular weight 35,000-50,000 kDa) was administered through the drinking water for 5 consecutive days (first round of DSS treatment). Subsequently, untreated water was given for 16 days. Then DSS was administered for 5 days followed by 16 days of untreated water (second round of DSS treatment). Finally, DSS was administered for 4 days (third round of DSS treatment) (Fig. 1A). Body weight was assessed at least 3 days per week throughout the course of the experiment.
Histopathological analysis. Mouse colon tissues were prepared as Swiss rolls fixed in 4% paraformaldehyde neutral buffer solution for paraffin embedding. Paraffin-embedded tissues were cut into 4 μm sections and stained with H&E. Histology scores were determined as described in a previous report 14 and were assigned based on the extent and severity of inflammation, ulceration, and hyperplasia of the mucosa in the distal colon.
Lamina propria cell isolation. Colons were opened longitudinally, cut into fine pieces, and incubated with Hanks' balanced salt solution (HBSS) containing 5 mM ethylenediamine tetraacetic acid for 15 min at 37 °C to remove epithelial cells. Colonic pieces without epithelial cells were then incubated with PBS containing 4% fetal Scientific RepoRts | 6:23920 | DOI: 10.1038/srep23920 bovine serum (FBS), 0.5 mg/mL collagenase type II, 1 mg/mL dispase, and 50 μg/mL DNase for 20 min at 37 °C. The cells were washed twice and suspended in PBS.
RT-PCR analysis. Total RNA was isolated using the RNeasy mini kit (Qiagen, Hilden, Germany) according to the manufacturer's recommended protocol. Total RNA from a small number of cells was isolated using RNAqueous ® -Micro Kit (Ambion, Austin, TX, USA). Residual genomic DNA was digested and removed using DNaseI (Invitrogen, Waltham, MA, USA). First-strand cDNA was synthesized using 1 μg total RNA and a High Capacity cDNA Reverse Transcription Kit (Applied BioSystems, Waltham, MA, USA) for RT-PCR. RT-PCR was performed using the Takara EX Taq (Takara, Shiga, Japan) according to the manufacturer's instructions. Real-time RT-PCR was performed using SYBR green (TOYOBO, Osaka, Japan) according to the manufacturer's instructions. Expression data were normalized to GAPDH mRNA expression. The primer sequences are shown in Supplementary Table 1.
Western blot analysis. Tissue samples were homogenized in standard radioimmunoprecipitation assay buffer with phenylmethylsulfonyl fluoride. Protein lysates were separated by 6-10% SDS-gel electrophoresis, and transferred to Immobilon-P membranes (Millipore, Darmstadt, Germany). Membranes were blocked in PBS with Tween buffer containing 3% skim milk for 1 hour at room temperature, probed with primary antibodies and secondary horseradish peroxidase-conjugated antibodies (GE Healthcare, Little Chalfont, UK), and developed using an enhanced chemiluminescence Western blot detection reagent (GE Healthcare).
16S rRNA gene-quantitative PCR analysis. 16S rRNA gene-quantitative PCR was performed as previously described 43 . Bacterial genomic DNA was isolated from fecal pellets using the QIAamp DNA Stool Mini Kit (Qiagen). Real-time PCR was performed using SYBR green (TOYOBO) according to the manufacturer's instructions. Data were normalized to those of total bacteria. The primer sequences are shown in Supplementary Table 1. ROS production analysis. Flow cytometry was used to measure intracellular ROS accumulation with 5-(and 6-) chloromethyl-2′, 7′-dichlorodihydrofluorescein diacetate, acetyl ester (CM-H2DCFDA) (Invitrogen). Isolated lamina propria cells were incubated in RPMI medium (Sigma-Aldrich) containing 5 μM CM-H2DCFDA for 30 min at 37 °C in the dark. Fluorescence was analyzed with a FACS Canto flow cytometer using FACS Diva software (BD Biosciences), and the data were analyzed with FlowJo software (TreeStar).
Statistical analysis. Data are expressed as mean ± standard error of the mean (SEM). The statistical analysis was carried out by one-way analysis of variance (ANOVA) followed by post-hoc analysis. P values less than 0.05 were considered statistically significant.