Fusobacterium nucleatum Increases Proliferation of Colorectal Cancer Cells and Tumor Development in Mice by Activating Toll-Like Receptor 4 Signaling to Nuclear Factor-κB, and Up-regulating Expression of MicroRNA-21

Abstract

Background & aims: Nearly 20% of the global cancer burden can be linked to infectious agents. Fusobacterium nucleatum promotes tumor formation by epithelial cells via unclear mechanisms. We aimed to identify microRNAs (miRNAs) induced by F nucleatum and evaluate their ability to promote colorectal carcinogenesis in mice.

Methods: Colorectal cancer (CRC) cell lines were incubated with F nucleatum or control reagents and analyzed in proliferation and would healing assays. HCT116, HT29, LoVo, and SW480 CRC cell lines were incubated with F nucleatum or phosphate-buffered saline (PBS [control]) and analyzed for miRNA expression patterns and in chromatin immunoprecipitation assays. Cells were incubated with miRNAs mimics, control sequences, or small interfering RNAs; expression of reporter constructs was measured in luciferase assays. CRC cells were incubated with F nucleatum or PBS and injected into BALB/C nude mice; growth of xenograft tumors was measured. C57BL adenomatous polyposis colimin/+, C57BL miR21a-/-, and C57BL mice with full-length miR21a (controls) were given F nucleatum by gavage; some mice were given azoxymethane and dextran sodium sulfate to induce colitis and colon tumors. Intestinal tissues were collected and tumors were counted. Serum samples from mice were analyzed for cytokine levels by enzyme-linked immunosorbent assay. We performed in situ hybridization analyses to detect enrichment of F nucleatum in CRC cells. Fusobacterium nucleatum DNA in 90 tumor and matched nontumor tissues from patients in China were explored for the expression correlation analysis; levels in 125 tumor tissues from patients in Japan were compared with their survival times.

Results: Fusobacterium nucleatum increased proliferation and invasive activities of CRC cell lines compared with control cells. CRC cell lines infected with F nucleatum formed larger tumors, more rapidly, in nude mice than uninfected cells. Adenomatous polyposis colimin/+ mice gavaged with F nucleatum developed significantly more colorectal tumors than mice given PBS and had shorter survival times. We found several inflammatory factors to be significantly increased in serum from mice given F nucleatum (interleukin 17F, interleukin 21, and interleukin 22, and MIP3A). We found 50 miRNAs to be significantly up-regulated and 52 miRNAs to be significantly down-regulated in CRCs incubated with F nucleatum vs PBS; levels of miR21 increased by the greatest amount (>4-fold). Inhibitors of miR21 prevented F nucleatum from inducing cell proliferation and invasion in culture. miR21a-/- mice had a later appearance of fecal blood and diarrhea after administration of azoxymethane and dextran sodium sulfate, and had longer survival times compared with control mice. The colorectum of miR21a-/- mice had fewer tumors, of smaller size, and the miR21a-/- mice survived longer than control mice. We found RASA1, which encodes an RAS GTPase, to be one of the target genes consistently down-regulated in cells that overexpressed miR21 and up-regulated in cells exposed to miR21 inhibitors. Infection of cells with F nucleatum increased expression of miR21 by activating Toll-like receptor 4 signaling to MYD88, leading to activation of the nuclear factor-κB. Levels of F nucleatum DNA and miR21 were increased in tumor tissues (and even more so in advanced tumor tissues) compared with non-tumor colon tissues from patients. Patients whose tumors had high amounts of F nucleatum DNA and miR21 had shorter survival times than patients whose tumors had lower amounts.

Conclusions: We found infection of CRC cells with F nucleatum to increase their proliferation, invasive activity, and ability to form xenograft tumors in mice. Fusobacterium nucleatum activates Toll-like receptor 4 signaling to MYD88, leading to activation of the nuclear factor-κB and increased expression of miR21; this miRNA reduces levels of the RAS GTPase RASA1. Patients with both high amount of tissue F nucleatum DNA and miR21 demonstrated a higher risk for poor outcomes.

Keywords: Carcinogenesis; Gene Regulation; Microbe; Signal Transduction.

Conflict of interest statement

Conflict of Interest: The authors have no competing financial, personal or professional interests to disclose.

Copyright © 2017 AGA Institute. Published by Elsevier Inc. All rights reserved.

Figures

Figure 1. F nucleatum promotes CRC cell proliferation and invasion in vitro and in vivo

A) HCT116 and LoVo cells were incubated with PBS, DH5α E coli or F nucleatum. The cell proliferation rates were evaluated by cell number counting after treatment at 6, 24 and 48 hr (*P < 0.05, ***P < 0.001, unpaired Student’s t test). B) The cell cycle distribution of treated or control cells was determined by flow cytometry-based assay after 48 hr (*P < 0.05, **P < 0.01, unpaired Student’s t test). C) The scratch wound healing assay was performed to evaluate the invasive capability of treated or control cells (F nucleatum treated cells vs control cells, *P < 0.05, **P < 0.01, unpaired Student’s t test). D) HCT116 cells treated with PBS, DH5α, or F nucleatum were subcutaneously injected into male BALB/C nude mice to produce xenograft tumors in animals (n=5 per group). The upper left panel of figures illustrate tumor growth curves (**P < 0.01 by one-way analysis of variance [ANOVA] and Bonferroni’s multiple comparison test) and tumor weight (**P < 0.01 by one-way ANOVA,). The right panel of figure shows the representative images of xenograft mice. The bottom left figures depict representative immunostaining (200× and 400×) for Ki-67 in xenograft tumor tissues. These results are representative of at least three independent experiments. Bars represent standard deviation (SD).

Figure 2. F nucleatum promotes tumorigenicity and CRC progression in ApcMin/+ mice

A) APCMin/+ mice were administrated with F nucleatum or PBS (negative control, NC) and sacrificed after treatment at 20 weeks. For the survival analysis, we compared survival time between F nucleatum treated group and PBS group. B) On 20 weeks, F nucleatum-treated APCMin/+ mice presented with ascites, bloody diarrhea, gut dilatation and splenomegaly. C) The left upper figure illustrates a representative image of a colon of APCMin/+ mice with or without F nucleatum treatment. The upper right figure and the bottom figure shows the administration of F nucleatum for 20 weeks which resulted in significant increase in tumor numbers, size and tumor load in the colorectum (n = 10 per group,*P<0.05, Mann-Whitney U test). D) The Kaplan-Meier survival curves show F nucleatum treatment reduced the survival rates of APCmin/+ mice (n = 8 per group, Log-rank (Mantel-Cox) test). E) Representative image of an immunostaining for PCNA in F nucleatum-treated or F nucleatum-untreated CRC tissues. F) The serum level of inflammatory factors in APCmin/+ mice with or without treatment of F nucleatum. (n = 10 per group, *P < 0.05, **P < 0.01, Mann-Whitney U test). Results represent means ± SD. The red arrows indicated the positive location.

Figure 3. F nucleatum regulates miR21 expression in CRC

A) Microarrays were performed to identify the differentially expressed miRNA between F nucleatum-treated and control cells. The miR21 expression was consistently up-regulated in different CRC cells after F nucleatum treatment (*P < 0.05, ***P < 0.001, unpaired Student’s t test). B and C) CCK-8 cell viability assay and wound healing assays illustrated that inhibition of miR21 completely abolished F nucleatum’s effect on cell growth and invasion(*P < 0.05, **P < 0.01 by unpaired Student’s t test). D) Both miR21 WT and knockout (KO) mice were initially administrated F nucleatum and then subjected to AOM/DSS treatment. At 20 weeks, mice were sacrificed for various experiments. E) The left representative image showed miR21 KO decreased the bloody diarrhea rates during the F nucleatum infection (n = 7 per group), and the right image showed a representative colon of the miR21 KO or wild type mice. F) The colorectum in miR21 KO mice had fewer tumor numbers, smaller size and less tumor load compared to WT mice (n = 7 per group,*P<0.05, Mann-Whitney U test). Furthermore, miR21 KO mice had longer survival rates than WT mice (n = 10 per group by log-rank (Mantel-Cox) test). These results are representative of at least three independent experiments. Results represent means values ± SD. The red arrows indicated the positively stained cells.

Figure 4. F nucleatum regulates expression of miR21 target gene RASA1 and activates MAPK signaling pathway

A) RASA1 was predicted as a putative miR21 target through analysis of 8 public prediction databases. B) The representative images of Western blot show overexpression or inhibiton of miR21 downregulated or upregulated RASA1 expression in HCT116 cells (n = 3 per group) and LoVo cells (n = 3 per group). C) Western blot showed RASA1 expression was lower in cancer tissues compared to matched normal tissues from CRC patients (n = 5 per group). D) We generated luciferase reporter plasmids which harbor either wild type (WT) or mutant (MT) miR21 binding sites in 3′-URT of RASA1. E) HCT116 and LoVo were transiently transfected with luciferase constructs along with miR21 mimics, inhibitors or negative controls. After 48hr, the luciferase activity was measured. All bars represent the mean values ± s.d. of three independent experiments. (*P < 0.05, **P < 0.01 by unpaired Student’s t test) F) The protein expression levels of RASA1 were significantly down-regulated in F nucleatum-infected CRC cells, while F nucleatum infections failed to increase RASA1 expression in normal intestinal epithelial cells NCM460 and IEC cells (n = 3). Compared to miR21 KO mice, the expression of RASA1 and another known miR21 target PDCD4 was significantly reduced, but P-ERK overexpressed in F nucleatum-infected miR21 WT mice tissues (n = 4 mice per group).

Figure 5. TLR4/MYD88/NFκB pathway is activated by F nucleatum infection

A) We treated HCT116 cells with F nucleatum or PBS (n = 3 per group) and then performed gene expression microarray analysis. The pathway analysis (KEGG and Reactome pathways) implicates F nucleatum infection can significantly stimulate TLR4/MYD88/NFκB pathway in CRC cells. B) HCT116 cells were incubated with F nucleatum, DH5α or PBS for 6h, 12h and 24h. NFκB subunit p65 and p50, phosphorylation level of p65 and p50 (P-p65 and P-p50), and inhibitor IκB-α were measured by Western blot. C) Silencing of NFκB significantly impaired the F nucleatum’s oncogenic effect on cell proliferation (proliferation assay) and cell invasion (wound healing assay) (*P < 0.05, **P < 0.01 by unpaired Student’s t-test. Bars represent SD of three experiments). D) qPCR results of F nucleatum DNA level in the low or high group (left figure, n = 6). Western blots analysis indicated that CRC tissues with high burden of F nucleatum showed activation of NFκB (right figure, n = 3). E)F nucleatum treatment induced the TLR4, TLR2, MYD88 expression in HCT116 (**P < 0.01, ***P < 0.001 by unpaired Student’s t-test. Bars represent SD of three experiments). F) qRT-PCR showed successfully silencing of TLR4 or MYD88 by Lv-TLR4 or Lv-MYD88 (**P < 0.01, ***P < 0.001 by unpaired Student’s t-test. Bars represent SD of three experiments). Western blots illustrated either TLR4 or MYD88 knockdown prevented NFκB activation from F nucleatum infection.

Figure 6. F nucleatum regulates miR21 expression through TLR4/MYD88/NFκB pathway

A and B) Knockdown of p65 in HCT116 or LoVo significantly down-regulated miR21 expression (**P < 0.01 by unpaired Student’s t test. Bars represent SD of three experiments). C) We generated luciferase reporter plasmids containing either wild type (WT) or mutant type (MT) of p65 binding sites in miR21 promoter. HEK293T cells were transfected with p65 shRNA or luciferase constructs (pGL3-WT and pGL3-MT) (**P < 0.01 by unpaired Student’s t test). D and E) HCT116 cells and LoVo were treated with F nucleatum or PBS following ChIP assay. qPCR results showed a more than 5 fold enrichment of miR21 promoter in p65 pulled-down DNA samples compared to IgG samples (**P < 0.01 by unpaired Student’s t test). F and G) The qRT-PCR assay indicated that F nucleatum failed to up-regulate miR21 expression when TLR4 or MYD88 was silenced in HCT116 and LoVo (**P < 0.01 by unpaired Student’s t test). H) Western blot analysis showed that CRC tissues from F nucleatum-treated APCmin/+ mice showed activation of NFκB, as well as other members including RASA1, PDCD4 and MAPK pathway (n = 4 mice per group).

Figure 7. Overabundance of F nucleatum correlates with high expression of miR21 in CRC and indicates poor clinical outcome

A) FISH assay showed F nucleatum (1000×) was enriched in the mucosa of CRC tissues and accompanied by high level of miR21 (1000×). The white arrows indicate positive staining. B) The qRT-PCR analysis showed that both the expression levels of F nucleatum and miR21 were higher in the cancer tissues than in the adjacent normal tissues (n = 90. **P < 0.01, ***P < 0.001 by unpaired Student’s t test. Bars represent SD). C) The qRT-PCR analysis showed F nucleatum and miR21 was overexpressed in cancer tissues in a stage dependent manner (n = 90. *P < 0.05, ***P < 0.001 by one way ANOVA and Bonferroni’s multiple comparison test. Bars represent SD). D) The amount of F nucleatum DNA was positively associated with miR21 expression in cancer tissues (n = 90. ***P < 0.001 by two-tailed nonparametric Spearman correlation.). E) Kaplan–Meier survival curve for 125 clinical specimens showed high amount of F nucleatum were associated with remarkably poor overall survival (OS) in CRC patients (log-rank (Mantel-Cox) test). Furthermore, subgroup of CRC patients with both high F nucleatum DNA level and miR21 expression has a higher risk clinical outcome (log-rank (Mantel-Cox) test). Neg, negative; L, low; H, high. F) The illustration of the hypothetical mechanism by which F nucleatum regulates miR21 expression through TLR4/MYD88/NFκB pathway in CRC.