Fusobacterium nucleatum Facilitates M2 Macrophage Polarization and Colorectal Carcinoma Progression by Activating TLR4/NF-κB/S100A9 Cascade

Fusobacterium nucleatum (Fn) has been considered as a significant contributor in promoting colorectal carcinoma (CRC) development by suppressing host anti-tumor immunity. Recent studies demonstrated that the aggregation of M2 macrophage (Mφ) was involved in CRC progress driven by Fn infection. However, the underlying molecular mechanisms are poorly characterized. Here, we investigated the role of Fn in Mφ polarization as well as its effect on CRC malignancy. Fn infection facilitated differentiation of Mφ into the M2-like Mφ phenotype by in vitro study. Histological observation from Fn-positive CRC tissues confirmed the abundance of tumor-infiltrating M2-like Mφ. Fn-induced M2-like Mφ polarization was weakened once inhibiting a highly expressed damage-associated molecular pattern (DAMP) molecule S100A9 mainly derived from Fn-challenged Mφ and CRC cells. In addition, Fn-challenged M2-like Mφ conferred CRC cells a more malignant phenotype, showing stronger proliferation and migration characteristics in vitro and significantly enhanced tumor growth in vivo, all of which were partially inhibited when S100A9 was lost. Mechanistic studies further demonstrated that activation of TLR4/NF-κB signaling pathway mediated Fn-induced S100A9 expression and subsequent M2-like Mφ activation. Collectively, these findings indicate that elevated S100A9 in Fn-infected CRC microenvironment participates in M2-like Mφ polarization, thereby facilitating CRC malignancy. Furthermore, targeting TLR4/NF-κB/S100A9 cascade may serve as promising immunotherapeutic strategy for Fn-associated CRC.


INTRODUCTION
Colorectal cancer (CRC) is a common digestive system neoplasm, ranking the third most common cancer worldwide and the fourth most frequent cause of cancer death following lung, liver, and stomach cancer (1). Over the past two decades, although early screening and detection have significantly reduced the incidence and mortality of adults aged 50 years with CRC, CRC still maintains a steadily rising trend in younger individuals (2,3). Recently, accumulating evidence by metagenome-wide association studies indicated an association between gut microbiota dysbiosis and CRC, and enteropathogenic microorganism plays a role in shaping the inflammatory environment and promoting tumor initiation and progression (4,5). Nevertheless, some key pathogenic bacteria as well as their promoting effect on CRC still need to be further clarified.
Fusobacterium nucleatum (Fn), a Gram-negative oral commensal anaerobe, has been found highly enriched in inflammatory bowel disease (IBD) and was regarded as a potential risk indicator of inflammation-associated CRC (6,7). Colonization of Fn contributed to carcinogenesis and was associated with the poor prognosis and short-term survival of clinical CRC patients (8)(9)(10). Previous studies have reported a direct regulating effect of Fn on human multi-type immunocyte, including lymphocytes and NK cells, facilitating periodontitis severity (11,12). In CRC, involvement of Fn in mediating tumor immune escape occurrence by regulating NK cell cytotoxicity and tumor-infiltrating T lymphocyte cell activities has also been reported (13). Recently, Fn infection was reported to increase tumor-associated macrophage (TAM) infiltration and participated in mediating M2-Mj polarization, facilitating CRC progression, suggesting a crucial antitumor effect of tumor-associated macrophages elicited by Fn infection (8). Nevertheless, detailed mechanistic investigations about the presence and activation state of Mj within Fn-infected CRC are scarce.
Inflammatory mediator S100A9 was secreted as a damageassociated molecular pattern (DAMP) in the tumor microenvironment (TME) of many inflammatory tumors, including gastric cancer, prostate cancer, and liver cancer (14)(15)(16). By binding its main TLR4 and RAGE receptors, S100A9 participated in regulating inflammatory-immune response and tumor progression (17). We previously found that elevated S100A9 in TME promotes CRC development by directly acting on the CRC cells and indirectly regulating myeloid-derived suppressor cell (MDSC)-mediated T cell immunosuppression (18,19). However, the regulatory mechanism of high levels of S100A9 in TME and whether S100A9 acts on other types of cells in the microenvironment remain elusive.
Recently, S100A9 has been reported to be regulated by some pathogen infection and exerted a promoting role in several types of cancer such as Hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC) and HPV-infected skin lesions and cancer (16,20). It is well-established that TLR4, an important receptor for pathogen-associated molecular pattern (PAMP), mediated downstream NF-kB activation in Fn-infected CRC cells, and was involved in the oncogenic cascade of CRC (21). S100A9, a newly identified NF-kB target gene identified in multi-type of cancer, was also demonstrated to be involved in myeloid cell differentiation into M2-like Mj in CRC (22,23). Thus, these observations led us to explore whether gut microbe Fn is involved in Mj activation by inducing the NF-kB/S100A9 cascade in a TLR4-dependent manner in Fn-infected CRC.
In the current study, we explored the impact of Fn on Mj polarization in the TME of CRC and the consequences on tumor growth in subcutaneous tumor formation in nude mice. We found that activation of TLR4-mediated NF-kB/S100A9 signaling pathways is involved in Fn-induced M2-like Mj polarization, which further promotes the progression of CRC. Our findings highlight the significance of S100A9 in regulating M2-like Mj polarization in Fn-infected CRC and implicate that S100A9 may serve as a potential intervention target in Fnassociated CRC patients.

Bacterial Strains and Culture Methods
Fn strain ATCC 25586 was purchased from the American Type Culture Collection (ATCC; Manassas, VA), and bacteria were cultured in brain heart infusion (BHI) containing hemin, K 2 HPO 4 , Vitamin K1, and L-Cysteine under anaerobic conditions at 37°C as previously described (24). E. coli was cultured for 24 h in Luria-Bertani (LB) medium overnight at 37°C in an orbital shaker incubator. Both bacteria were collected by 4°C centrifugation at 6,000×g for 8 min, and bacterial suspension was prepared after washing three times with sterile PBS. Turbidimetry was used to determine the turbidity of bacterial suspensions and adjust its concentration to 1 × 10 8 CFU/ml. The following experiments were all processed at a multiplicity of infection (MOI) of 100:1.

Cell Line Culture and Mj Differentiation of THP-1
Human CRC cell lines HCT116 and SW480 and human monocyte THP-1 were presented by the Chongqing Key Laboratory of Molecular Oncology and Epigenetics, the First Affiliated Hospital of Chongqing Medical University. HCT116 and SW480 cells were cultured in Dulbecco's modified Eagle's medium (DMEM), and THP-1 cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium. Both media were supplemented with 10% fetal bovine serum (FBS) (Gibco, USA). The final concentrations of penicillin and streptomycin (Hyclone, USA) in these two media were 100 U/ml and 100 µg/ ml, respectively. Cell culture was maintained at 37°C in an incubator containing 5% CO 2 , 95% air.
THP-1 cells were treated with PMA (50 ng/ml) (Sigma, USA) for 24 h to obtain differentiated Mj as demonstrated previously (25). Successfully differentiated Mjs were washed twice with PBS and incubated in fresh medium to be used for the following experiments.

Clinical Specimens
CRC and matching distal normal tissues were collected from 16 patients who had undergone colorectal resection at the First Affiliated Hospital of the Chongqing Medical University. The clinicopathological data of the subjects including gender, age, Dukes staging, and lymphatic metastasis at initial diagnosis are shown in Table S1. The patients received no chemotherapy, hormonal therapy, or radiotherapy before surgery, and written informed consent was received from all participants. This study was approved by the Ethics Committee of Chongqing Medical University (protocol number 2012-19).

Preparation of the Recombinant Proteins
Recombinant S100A9 (rS100A9) proteins used in this study have been described previously (18). Briefly, the pGST-moluc and pGST-moluc-S100A9 were transformed into BL21 bacteria following the instructions for calcium chloride transformation. Isopropylthio-b-D-galactoside was used to induce the expression of GST and GST-S100A9 proteins. The supernatant of sonicated bacteria was collected, and then incubated with glutathionesepharose 4B beads. Then, rS100A9 and control GST on the beads were eluted using elution buffer with reduced glutathione. Finally, the two recombinant proteins were filtered with 0.22 mm membrane and stored at −80°C.

Preparation of Conditioned Medium From Cells With Various Treatments
Mj or CRC cells were seeded in six-well plates and then infected with bacteria. After 48 h, the supernatant was harvested and clarified by centrifugation to remove the bacteria, cells and their debris, and then stored at −80°C refrigerator for later use.
Mj or CRC cells were transfected with non-specific siRNA (siNC) or S100A9-siRNA (siS100A9) (GenePharma, China) ( Table 1) at a final concentration of 100 pM with lipofectamine 2000 reagent (Invitrogen, USA) in serum-free medium. The medium was replaced with fresh medium containing 10% FBS after 6 h of transfection and Fn was added to each group. The supernatant was collected after 48 h in the same manner as before.
Mj or CRC cells were treated with TAK-242 (5 µM) and BAY 11-7082 (10 µM) for 1 h prior to Fn stimulation. The supernatant was collected after 48 h in the same manner as before.
Mj cells were treated with GST (10 µg/ml) or rS100A9 (10 µg/ml). After 24 h, we changed the media with fresh media and collected the supernatant after incubating for another 48 h in the same manner as before.
ALL mentioned conditioned medium (CM) was finally made up by mixing the supernatant with 10% FBS-supplemented fresh medium in a 1:1 ratio to make 5% FBS-containing CM.

ELISA
Concentration of S100A9, TNF-a, and IL-10 in the abovementioned Conditioned media (CM) with different treatments was measured with specific ELISA kits (JYM, China) according to the manufacturer's instructions.

Cell Proliferation Assay
Cell proliferation was assessed by CCK-8 assay using a Cell Counting Kit (Dojindo, Japan) following the manufacturer's protocol. Briefly, HCT116 and SW480 cells were seeded in 96well plates at 3,000 cells/well and cultured for described hours.
Then CCK8 solution was added. Finally, the absorbance at 450 nm of each well was measured daily using a microplate reader. Each condition was done in quintuplicate, and the experiment was repeated thrice.

Cell Migration Assay
Cell migration assay was examined to assess the cell migratory capacity of CRC cells using 24-well Transwell cell culture chambers (pore size, 8 mm). Briefly, CRC cells were seeded at a density of 2 × 10 4 cells per well in serum-free medium at the upper chamber, and the corresponding CM was added in the lower chamber to establish co-culture systems. After incubating at 37°C for 24 h, the transmembrane cells were dried, fixed with methanol, and stained with commercialized crystal violet staining solution (Beyotime, China) and the migration cell number in five to eight fields of each insert was counted under an inverted microscope at a magnification of ×100. The experiment was repeated thrice.

RNA Isolation and Quantitative Real-Time PCR
Total RNA extraction from the treated cells or tumor tissues was performed with TRIzol reagents (Invitrogen). cDNA samples were synthesized using random primers from 1 mg total RNA with a Reverse Transcription kit (Takara, Japan). The mRNA levels of S100A9, iNOS, TNF-a, IL-10, CD206, E-cadherin, N-cadherin, Vimentin, TGF-b, VEGF, and TLR4 were analyzed and normalized to the GAPDH with the CFX96 real-time PCR detection system (Bio-Rad, USA) using SYBR Green dye (Biomake, China) according to the manufacturer's instructions. The primers in this study were synthesized by Genscript, and the primer sequence information is shown in Table 1.

Western Blot
The cells of different treatment groups were collected, and total cellular protein was extracted with RIPA buffer containing phosphatase/protease inhibitor after washing three times with ice-cold PBS. The cell lysates were collected after centrifugation, and the concentration of protein was determined by BCA assay. The extraction of xenograft tumor protein was carried out with reference to the kit instructions (BestBio, China) according to the manufacturer's instructions. Samples containing equal amount of proteins were separated using SDS-PAGE according to the molecular weight of different proteins and transferred to PVDF membranes. Then the membranes were blocked with 5% non-fat dry milk and incubated with anti-S100A9 (Abcam, UK), anti-CD86 (Bioss, China), anti-CD206 (Proteintech, China), anti-total NF-kB p65 (Cell Signaling Technology, USA), anti-Phospho-NF-kB p65 (p-p65) (Ser536) (Cell Signaling Technology, USA), anti-Ecadherin (Immunoway Biotechnology, USA), anti-N-cadherin (Cell Signaling Technology, USA), anti-PCNA (Wanleibio, China), anti-VEFG (Santa Cruz Biotechnology, USA), anti-TGFb (Abcam, UK) or anti-b-actin (Cell Signaling Technology, USA), followed by incubation with secondary antibodies conjugated with horseradish peroxidase. The proteins of interest were detected using the SuperSignal West Pico Chemiluminescent Substrate kit. The results were recorded by Bio-Rad Electrophoresis Documentation (Gel Doc 1000, Bio-Rad, USA) and Quantity One Version 4.5.0.

Flow Cytometry
To investigate Mj differentiation, M1 marker (CD86) and M2 marker (CD206) in different treatment groups were analyzed by flow cytometry. Briefly, the prepared cells were collected and washed with PBS and then stained with PE-mouse anti-human CD86 (BioLegend, USA) or PE-mouse anti-human CD206 (BioLegend, USA) for 30 min at 4°C. Cells were then washed and analyzed using FACSVantage SE flow cytometer (Becton-Dickinson, USA).

Immunofluorescence
Mjs on the climbing piece in 24-well culture plates were treated with or without bacterium infection or treated with relevant CM for 48 h. Then the cells were fixed with 4% paraformaldehyde for 20 min, washed with PBS, and permeabilized with 0.01% Triton X-100 for 10 min. After being washed with PBS, the cells were blocked with 10% goat serum for 30 min at room temperature and incubated with anti-CD86 monoclonal antibodies (Santa Cruz, USA) and anti-CD206 antibodies (Proteintech, China) at 4°C overnight. The next day, the cells were rinsed with PBS for clearing the primary antibody, and then incubated with Alexa Fluor 488-conjugated goat anti-mouse secondary antibody (Beyotime, China) or Cy3conjugated goat anti-rabbit secondary antibody (Beyotime, China) at room temperature for 1 h in the dark, and then washed with PBS, and counterstained the nucleus with DAPI for 10 min. After washing again with PBS and mounting with antifade polyvinylpyrrolidone mounting medium (Beyotime, China), the fluorescent images were observed using confocal microscope (Lecia, Germany).

Xenograft
Four-week-old female BALB/c nude mice were purchased from Beijing Huafukang Biotechnology. For the xenograft experiments, mice were randomly divided into four groups (n = 3 in each group). HCT116 cells were mixed with equal proportion of Mj treated with control E. coli, Fn, Fn + siNC or Fn + siS100A9 for 48 h and suspended in 100 µl PBS for the subsequent use. Then the mixed cells (1 × 10 7 ) in 100 µl PBS were injected subcutaneously into the posterior flank of each nude mouse to establish the xenograft model. The length and width of a tumor were detected using a caliper every three days. Tumor volumes were calculated according to the following formula: Volume = (width) 2 × length/2. The mice were sacrificed after 21 days, and the tumor tissues were removed, weighed, fixed in 4% buffered formaldehyde, embedded in paraffin, and sectioned for further immunohistochemical analysis. All the in vivo experiments were approved and conducted in accordance with the guidelines established by the University Animal Care and Use Committee for Laboratory Animal Research in Chongqing Medical University (protocol number 2018-003).

Microbial Fluorescence In Situ Hybridization Analysis
The analysis for detecting invasive Fn in CRC tissue used Fntargeted probe, FUS664 (FITC-labeled), 5′-CTT GTA GTT CCG C(C/T) TAC CTC-3′ (26). Paraffin-embedded slides were prepared and hybridized with reference to previously described methods (26). Bacteria were counted in five random fields per slide at a magnification of ×100 by an observer blind to the sample status, and the number of bacteria per field was calculated on average. The average number of Fn in each field in the case of <5, between 5 and 20, and >20 visualized FUS664 probes was defined as negative, weak, and positive Fn abundance, respectively.

Immunohistochemical Procedures
Indirect immunohistochemistry (IHC) analysis of formalin-fixed and paraffin-embedded tissue sections was performed. In brief, the sections were deparaffinized, dehydrated, boiled in 0.01 M citrate buffer for 10 min, and then incubated with 0.3% H 2 O 2 in methanol for 10 min to block endogenous peroxidase activity. The sections were incubated with anti-S100A9 (Abcam, UK), anti-CD68 (Santa Cruz, USA), anti-CD86 (Santa Cruz, USA), anti-CD206 (Proteintech, China), anti-E-cadherin (ImmunoWay Biotechnology, USA), and anti-N-cadherin (Cell Signaling Technology, USA) following incubation with secondary antibody tagged with the peroxidase enzyme for 30 min at room temperature and were visualized with 0.05% DAB till the desired brown reaction product was obtained. All slides were counterstained with hematoxylin and then observed using Nikon E400 Light Microscope, and representative photographs were taken. For double immunofluorescence staining, the sections were incubated with anti-S100A9 together with anti-CD68 (HUABIO, China), anti-CD86 (Santa Cruz, USA), anti-CD206 (Proteintech, China), or anti-cytokeratin 20 (Proteintech, China) antibodies, followed by incubation with secondary antibodies Alexa fluor 647-conjugated anti-mouse IgG and Alexa fluor 488-conjugated anti-rabbit IgG. These sections were also stained with 10 mg/ml DAPI. The fluorescent images were then observed and analyzed using a multi-laser confocal microscope.

Statistical Analysis
Data from the two groups were analyzed with the two-tailed Student's t-test and data from three or more groups using oneway ANOVA followed by Newman-Keuls' multiple comparison test. All experiments were independently repeated three times. The statistical analyses were carried out using SPSS version 13.0. Statistical differences are presented at probability levels of p <0.05, p <0.01, and p <0.001.

Involvement of Fn in M2-Like Polarization of Mj
To investigate whether the phenotypic differentiation of Mj is affected by Fn infection in vitro, M1 markers (inducible nitric oxide synthase, iNOS; tumor necrosis factor, TNF-a; the B7related cell surface proteins B7-2, CD86) and M2 markers (interleukin-10, IL-10; the mannose receptor, MRC/CD206) were determined. Compared to the control E. coli-challenged Mj, Fn-challenged Mj harbored significantly lower mRNA level of M1 markers iNOS and TNF-a but significantly higher mRNA level of M2 markers IL-10 and CD206 (Figures 1A, B). Similar tendency for protein levels of TNF-a and IL-10 in the supernatants was confirmed by ELISA ( Figure S1). These results suggested that Fn may be involved in M2-like polarization. Marked CD206 expression but less CD86 expression was detected in Fn-challenged Mj by western blot (Figure 1C). High intensity of the signal for CD206 but low intensity of signal for CD86 was also confirmed by immunofluorescence and flow cytometry analyses ( Figures  1D-F). Collectively, these findings indicated that Fn might contribute to M2-like Mj differentiation.

Elevated S100A9 Expression in Fn-Challenged Mj and CRC Cells
Increased level of S100A9 is related to the infection of multiple pathogens and participates in disease progression by regulating inflammatory immunity (27,28). Therefore, we analyzed the S100A9 level in Fn-challenged Mj and CRC cells. The mRNA level of S100A9 was significantly up-regulated in the Fnchallenged Mj (Figure 2A). Consistently, its protein level in Fn-challenged Mj as well as the supernatant was also markedly increased compared to the control groups ( Figures 2B, C). In line with the finding, Fn-infected CRC cells as well as the supernatant also showed high level of S100A9 than the control group ( Figures 2D-F). These data suggested that S100A9 may be elevated in the microenvironment of CRC with Fn infection. Next, we evaluated the enrichment of Fn and detected CD68 + Mj, CD86 + M1-like Mj, CD206 + M2-like Mj, and S100A9 expression in CRC tissues. As expected, there was more Mj infiltration in Fn-positive CRC tissues compared to Fn-negative CRC tissues, and most of them were M2-like Mjs accompanied by higher levels of S100A9 in the TME (Figures 2G and S2). The results suggested that S100A9 expression is elevated in Fninfected Mj and CRC cells, which may be associated with M2like Mj polarization in CRC tissues.

S100A9 Is Involved in M2-Like Mj Polarization in the Inflammatory Microenvironment of CRC With Fn Infection
The results above showed that S100A9 was highly expressed in Fn-infected CRC, but the direct evidence about the relation of higher S100A9 and M2-like Mj polarization needs to be further studied. Therefore, we analyzed the Mj polarization status in siS100A9-transfected Mj cultured with Fn. As expected, S100A9 expression was successfully reduced in the supernatant of siS100A9-transfected Mj with Fn infection ( Figure S3). Knockdown of S100A9 expression obviously blocked Fninduced increase of mRNA levels of M2 markers IL-10 and CD206 but improved Fn-induced decrease of mRNA levels of M1 marker iNOS and TNF-a ( Figures 3A, B). The protein levels of TNF-a and IL-10 in the supernatants were confirmed by ELISA ( Figure S4). Consistently, down-regulated M2 marker CD206 and up-regulated M1 marker CD86 were observed in siS100A9transfected Mj infection with Fn by western blot ( Figure 3C). Low intensity of signal for CD206 but high signal for CD86 was also confirmed in siS100A9-transfected Mj infection with Fn by immunofluorescence assay (Figure 3D). To better illustrate the effect of S100A9 on M2-like Mj polarization in the Fn-infected CRC inflammation microenvironment, CRC cells were transfected with siS100A9 followed by Fn infection, and then CM was collected for further culturing with Mj ( Figure 3E). siS100A9 transfection significantly inhibited the secretion of S100A9 protein in Fn-infected CRC cells compared to the control groups ( Figure 3F). Moreover, Mj tended to be polarized towards M1-like Mj (mRNA levels: iNOS high , TNFa high , IL-10 low , CD206 low ; protein levels: CD86 high , CD206 low ) after culturing with the CM from CRC cells transfected with siS100A9 followed by Fn infection (Figures 3G-I). In contrast, after treatment with rS100A9 at a final concentration of 10 mM for 48 h, Mjs were apt to be M2 activated (mRNA levels: iNOS low , TNF-a low ; IL-10 high , CD206 high ; protein levels: CD86 low , CD206 high ) ( Figures 3J-L). In conclusion, these results directly supported that elevated S100A9 favors the M2 polarization of Mj in the TME of Fn-infected CRC.

Fn-Challenged M2-like Mj, Which Is Mediated by S100A9, Promotes the Proliferation and Migration of CRC Cells
To further investigate the effects of Fn-challenged Mj on the proliferation and migration ability of HCT116 and SW480 cells, CCK8 and Transwell assays were conducted after CRC cells were treated with various Mj CM. There was a significant increase in the proliferation (Figures 4A, B) and migration ( Figure 4C Figures 4D-F). The remarkably enhanced ability of proliferation ( Figures 4G, H) and migration ( Figure 4I) of HCT116 and SW480 cells was confirmed again when these cells were cultured with rS100A9treated Mj CM [(Mj + rS100A9)-CM). In addition, epithelialto-mesenchymal transition (EMT)-related genes N-cadherin and Vimentin were up-regulated, but E-cadherin was decreased in HCT116 cells cultured with (Mj + Fn)-CM ( Figure S5). These results, along with those aforementioned, suggested that Fnchallenged M2-like Mj, which was mediated by S100A9 in the TME, confers onto CRC cells a more malignant phenotype.
Fn Up-Regulates S100A9 Expression and Promotes M2-Like Polarization via the TLR4/NF-kB Pathway To further examine the mechanisms by which Fn engages in S100A9 expression and M2 polarization, we focused on TLR4/ NF-kB pathway, which is substantially activated in CRC with Fn enrichment (21) and determine whether it takes part in the effect. The mRNA levels of TLR4 in Mj, HCT116, and SW480 cells were obviously increased in response to Fn infection ( Figure 5A). Phospho-NF-kB p65 (p-p65) levels in Mj, HCT116, and SW480 cells were also increased with time and reached a peak almost at 120 min, while there was no obvious change in the total p65 ( Figure 5B). Fn-induced p-p65 in the three cell lines was inhibited after TLR4 inhibitor TAK-242 stimulation ( Figure 5C), suggesting TLR4 mediated Fn-mediated NF-kB activation. Especially, treatment with TLR4 inhibitor TAK-242 and NF-kB inhibitor BAY 11-7082 also inhibited S100A9 protein levels in Fnchallenged Mj and CRC cells ( Figure 5D) as well as its levels in cell supernatant (Figures 5E-G). Exposure of Fn-challenged Mj to the TAK-242 or BAY 11-7082 failed to induce M2-like Mj (mRNA levels: iNOS high , TNF-a high , IL-10 low , CD206 low ) ( Figures  5H, I). Consistently, this change in Mj phenotype was again proved when cultured with the CM from Fn-challenged CRC cells, which were pretreated with TAK-242 and BAY 11-7082 ( Figures  5J, K). Altogether, these results indicated that the TLR4-mediated NF-kB pathway is partially responsible for M2-like polarization induced by S100A9 in response to Fn infection in CRC.

Accelerated Growth of the Subcutaneous Tumor by M2-Like Mj Induced by Fn
In the xenograft nude mouse models, HCT116 and differently treated Mj were inoculated into nude mice. The results from the  A, B, D, E, G, H), data shown are mean absorbances ± SD. In (C, F, I), data shown are mean migrating cells ± SD. ns, not significant. *p < 0.05, **p < 0.01, ***p < 0.001. nude mouse xenograft model showed that co-injection with HCT116/Fn-treated Mj significantly accelerated the tumor growth, and this effect was partially blocked by silencing S100A9 expression in Mj in the nude mice co-injected with HCT116/(Fn + siS1A00A9)-treated Mj ( Figure 6A). Tumor growth curve and the average weight of each group also showed the same tendency ( Figures 6B, C). In addition, IHC results showed that the protein levels of proliferation-related marker PCNA, EMT-related marker E-cadherin and N-cadherin, and tumor-promoting molecules VEGF and TGF-b were higher in tumors of nude mice co-injected with HCT116/Fn-treated Mj than those of the compared control groups, while these protein levels were decreased in the nude mice co-injected with HCT116/ (Fn + siS1A00A9)-treated Mj ( Figure 6D). Western blot and qPCR results about the levels of PCNA, E-cadherin, N-cadherin, VEGF, TGF-b in each xenograft tumor were also consistent with the result of IHC (Figures 6E-I). These data suggested that the interaction between HCT116 and Fn-treated Mj promotes tumor growth in nude mice in vivo, which may be involved in Fn-induced M2-like polarization mediated by S100A9.

DISCUSSION
Identifying specific carcinogenic microorganisms remains the focus of CRC research (29). Recently, among these identified and published CRC-related gut microbes, Fn has attracted the most attention. Although involvement of Fn in immunosuppressive TME has been indicated, the detailed mechanisms by which Fn participated in regulating the activation of immune cells have not been well-elucidated. Here, we demonstrated a regulatory role of Fn in CRC TME on M2-like polarization, which is mediated by activation of TLR4-dependent NF-kB/S100A9 signaling pathway in Mjs and CRC cells (Figure 7). Mjs are dynamic and heterogeneous cells whose differentiation, tissue distribution, and responsiveness to stimuli are governed by different mechanisms (30). Based on their response to different microenvironmental signaling factors, Mjs can be differentiated into two classifications, namely, M1-like Mj sand M2-like Mjs. M1-like Mj mainly plays a pro-inflammatory and anti-tumor role, while M2-like Mj is often considered to be tumor-promoting (31). Viral, parasitic, fungal, and bacterial pathogens can induce M2 activation in Mj and further promote pathogen-induced inflammatory injury and even canceration (32). Numerous studies have elucidated Fn, a Gram-negative oral commensal anaerobe, which is one of the most prevalent species in oral inflammation diseases and extraoral infections (33,34). At the same time, Fn has been reported to be highly enriched in CRC tissues and helps tumor cells to achieve immune evasion by inhibiting immune cell activity in TME (13,(35)(36)(37). However, the presence and activation state of Mj in Fn-infected CRC remain to be elucidated. In this study, we determined that Fn-challenged Mjs were inclined to M2-like activation, which in turn enhances the proliferation and migration of CRC cells in vitro. Furthermore, our results from the xenograft nude mouse also supported a positive association between Fn-challenged M2-like Mj and tumor growth. Altogether, our results suggest that Fn-induced M2-like polarization facilitated CRC progression. Compelling links are beginning to appear among bacteria, immune escape, and CRC. Base on Fn abundance in CRC along with weaken host anti-tumor immunity, there is a great need for further mechanistic understanding of CRC to find a novel target molecule. S100A9, originally known as immunogenic protein, is mainly expressed by bone marrow-derived cells such as mononuclear, Mj, and neutrophils (17). Besides, S100A9 induces positive feedback and promotes further leukocyte recruitment under pathological conditions associated with inflammatory infection (38). Experimental data indicated that S100A9 presents abundant expression in various inflammation-associated human cancers, no doubt including CRC (17,18). Our previous work confirmed elevated S100A9 levels in CRC, stimulating survival and migration of CRC cells (18). Here, we further determined that high S100A9 levels in CRC may be associated with Fn infection. Moreover, the number of tumor-infiltrating M2-like Mj was more in Fn-positive CRC tissues accompanied by higher S100A9 expression than that in Fn-negative tissues with lower S100A9 expression. These results further prompt us to infer that S100A9 may play a potential role in regulating Mj toward M2-like phenotype. Previously, tumor-infiltrating S100A9-positive inflammatory or immune cell tissues were closely related to the pathological stage of CRC (22). S100A9 was also reported to be involved in THP-1 cells differentiated into MDSCs or M2-like Mj stimulated by myofibroblast-based CM (39). Here, our results demonstrated that up-regulation of S100A9 expression from Fninfected Mj and CRC cells in the TME was partially responsible for M2-like Mj activation state. Not surprisingly, the proliferation and migration ability of CRC cells co-cultured with Fn-treated Mj would be weakened by silencing S100A9 expression in the TME of CRC.
In this study, we attempted to uncover the mechanism for regulation of S100A9 expression and M2-like Mj polarization by Fn infection. As an inflammation/immunity mediator, high levels of S100A9 have been indicated and aggravated mutiple pathogen infections, including coxsackievirus B3 (CVB3)-induced myocarditis (27), streptococcus pneumoniae (SP)-related otitis media (OM) (40), and influenza A virus-related pneumonia (28). The TLR4/NF-kB signal is often activated in pathogen infection. Influenza A virus could up-regulate S100A9 expression and synergistically activate TLR4 signaling cascade in lung Mj and endothelial cells, exaggerating pro-inflammatory response, cell-death, and virus pathogenesis (28). Our previous study also found HBV-induced NF-kB activation enhances transcription of S100A9 by binding to its promoter, contributing to HCC malignancy (16). In human cytomegalovirus (HCMV)related inflammatory breast cancer, NF-kB activation could result in M2-like Mj polarization accompanied with the secretion of tumor-promoting M2-like cytokines (41,42). Thus, we speculated that Fn infection may facilitate CRC malignancy involving Mj activation, either directly or indirectly, via activating specific signal pathways such as that of TLR4-dependent NF-kB. Now we have shown that PAMP extracted from Fn and Fn itself could be recognized by TLR4, which further activates its downstream signaling cascade in CRC (8,21). Also, NF-kB, as a widely acknowledged downstream effector of TLR4 signaling, is highly activated in Fn-associated CRC (6). Interestingly, we found a significant increase in the expression of TLR4 and time-dependent activation of NF-kB caused by Fn infection in Mj and CRC cells, which was inhibited by TLR4 inhibitor TAK-242. In addition, S100A9 expression and M2 activation were also significantly suppressed when the TLR4/NF-kB pathway was blocked. All the FIGURE 7 | Schematic diagram of the relationship among Fn, M2 polarization, and CRC progression. Activation of the TLR4/NF-kB cascade in Mj and CRC cells by Fusobacterium nucleatum infection mediated high levels of S100A9 in CRC microenvironment, which led to M2-like Mj phenotype showing increased IL-10 and CD206 levels, contributing to CRC malignance. evidence above suggested that interference with TLR4/NF-kB/ S100A9 cascade should be given enough attention in the treatment of Fn-associated CRC.
In conclusion, the current findings demonstrate that S100A9 plays a critical role in regulating M2-like Mj polarization in Fninfected CRC microenvironment, and targeting TLR4/NF-kB/ S100A9 cascade may attenuate the immunosuppressive effect and serve as promising immunotherapy strategy for Fn-associated CRC.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors without undue reservation.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by the Ethics Committee of Chongqing Medical University. The patients/participants provided their written informed consent to participate in this study. The animal study was reviewed and approved by the Ethics Committee of Chongqing Medical University.

AUTHOR CONTRIBUTIONS
LH and YL performed the experiments, analyzed the data, and wrote the manuscript. XK, RW, QP, and YZ analyzed the data. LD and LZ conceived ideas, oversaw the research, and co-wrote the manuscript. All authors contributed to the article and approved the submitted version.