Fusobacterium nucleatum promotes colorectal cancer cells adhesion to endothelial cells and facilitates extravasation and metastasis by inducing ALPK1/NF-κB/ICAM1 axis

Abstract

Metastasis is the leading cause of death for colorectal cancer (CRC) patients, and the spreading tumor cells adhesion to endothelial cells is a critical step for extravasation and further distant metastasis. Previous studies have documented the important roles of gut microbiota-host interactions in the CRC malignancy, and Fusobacterium nucleatum (F. nucleatum) was reported to increase proliferation and invasive activities of CRC cells. However, the potential functions and underlying mechanisms of F. nucleatum in the interactions between CRC cells and endothelial cells and subsequent extravasation remain unclear. Here, we uncovered that F. nucleatum enhanced the adhesion of CRC cells to endothelial cells, promoted extravasation and metastasis by inducing ICAM1 expression. Mechanistically, we identified that F. nucleatum induced a new pattern recognition receptor ALPK1 to activate NF-κB pathway, resulting in the upregulation of ICAM1. Interestingly, the abundance of F. nucleatum in tumor tissues of CRC patients was positively associated with the expression levels of ALPK1 and ICAM1. Moreover, high expression of ALPK1 or ICAM1 was significantly associated with a shorter overall survival time of CRC patients. This study provides a new insight into the role of gut microbiota in engaging into the distant metastasis of CRC cells.

Keywords: ALPK1; Gut microbes; ICAM1; adhesion; colorectal cancer metastasis.

Figure 1.

F. nucleatum promotes CRC cells adhesion to endothelial cells and facilitates extravasation by upregulating ICAM1. (a) Schematic illustration of adhesion assay in vitro. (b) GFP-labeled HCT116 cells pretreated with F. nucleatum, E. coli or PBS control for 6 hours were subjected to adhesion assay. Representative images of adherent cells were shown (left) and the adherent cells were quantified by counting in five fields (right). Scale bar, 100 μm. (c) HCT116 cells were pretreated with F. nucleatum, E. coli or PBS control for 6 hours and subjected to migration assay. The migrated cells were quantified at 18 hours by counting in five fields. Scale bar, 100 μm. (d) Schematic illustration of cancer cell extravasation model in vivo. (e) The colonized GFP-labeled HCT116 cells in mice lung tissues were quantified by immunofluorescence assay, in which lung sections were stained with a marker of vascular endothelial cells CD31 (red), and the nuclei were counterstained with DAPI (blue). Scale bar, 100 μm. (f) 256 genes with significant mRNA upregulation (fold change > 2, p.adj < 0.05) in RNA-sequencing, 880 genes in the GO class of cell-cell adhesion (GO: 0098609), and 133 genes in the KEGG pathway of cell adhesion molecules (hsa04514) with key functions were subjected to Venn diagram analysis. Venn diagram showing the overlapping genes. (g) Western blot analysis was performed to detect the protein levels of ICAM1 in HCT116 cells and LoVo cells treated with F. nucleatum, E. coli or PBS control. (h) Quantitative RT-PCR analysis of ICAM1 in PDX tumor tissues treated with F. nucleatum or PBS control. Data are shown as mean ± SD. ** P< .01, *** P < .001, **** P < .0001, Student’s t test (b, c, h), Mann-Whitney test (e).

Figure 2.

ICAM1 is involved in F. nucleatum-induced CRC cells adhesion to endothelial cells and migration in vitro. (a-b) Quantitative RT-PCR (a) and Western blot analysis (b) of ICAM1 were performed in HCT116 cells transfected with two siRNAs targeting ICAM1 or control siRNAs. (c-d) Adhesion assay (c) and migration assay (d) of HCT116 cells transfected with the indicated siRNAs. The migrated cells were observed at 24 hours. Scale bar, 100 μm. (e-f) Quantitative RT-PCR (e) and Western blot analysis (f) were performed in HCT116 cells. They were transfected with the indicated siRNAs, and then co-cultured with F. nucleatum or PBS. (g-h) HCT116 cells transfected with the indicated siRNAs were co-cultured with F. nucleatum, and subjected to adhesion assay (g) and migration assay (h). The migrated cells were observed at 18 hours. Scale bar, 100 μm. Data are shown as mean ± SD. *** P < .001, **** P < .0001, by Student’s t test.

Figure 3.

ICAM1 is involved in F. nucleatum-mediated CRC cells extravasation and metastasis in vivo. (a) GFP-labeled HCT116 cancer cells transfected with two siRNAs targeting ICAM1 were co-cultured with F. nucleatum or PBS, and subjected to extravasation model in vivo. The colonized HCT116 cells in lung tissues were measured by immunofluorescence assay. Scale bar, 100 μm. (b-c) HCT116 cells stably infected with lentivirus-based ICAM1 shRNAs or control shRNAs were co-cultured with F. nucleatum or PBS control, and subjected to Western blot analysis (b) and migration assay (c). The migrated cells were quantified at 18 hours. Scale bar, 100 μm. (d-e) The ICAM1-knockdown HCT116 cells pretreated with F. nucleatum or PBS were tail-vein injected into nude mice to develop pulmonary metastases. Representative gross lungs and H&E stained lung sections were shown (d). Arrows indicated metastatic nodules. Pulmonary metastatic nodules per mice were quantified (e). Data are shown as mean ± SD. * P < .05, ** P < .01, **** P < .0001, Mann-Whitney test (a, e), Student’s t test (c).

Figure 4.

F.nucleatum upregulates ICAM1 expression through the activation of NF-κB signaling pathway. (a) KEGG pathway analysis of a total number of 503 genes upregulated with significant difference in RNA-sequencing. (b) GSEA showed the differentially expressed gene cluster related to NF-κB signaling pathway in HCT116 cells with or without F. nucleatum treatment. (c) Western blot analysis of phospho-IκBα, NF-κB subunit p65 and phospho-p65 in HCT116 cells treated with F. nucleatum, E. coli or PBS control. (d) Immunofluorescence assay of NF-κB p65 distribution in the indicated HCT116 cells. Cells were stained with specific antibody against p65 (green), and the nuclei were counterstained with DAPI (blue). Scale bar, 20 μm. (e-f) HCT116 cells were co-cultured with F. nucleatum or PBS control, and then treated with NF-κB inhibitor, BAY11-7082. Quantitative RT-PCR (e) and Western blot analysis (f) were performed. (g-h) HCT116 cells were transfected with two siRNAs targeting p65, and then co-cultured with F. nucleatum or PBS control. Quantitative RT-PCR (g) and Western blot analysis (h) were performed. (i-j) Adhesion assay (i) and migration assay (j) of HCT116 cells with indicated treatments. The migrated cells were observed at 18 hours. Scale bar, 100 μm. (k-l) P65-depleted HCT116 cells transfected with the indicated plasmids were applied for Western blot analysis (k) and adhesion assay (l). Scale bar, 100 μm. Data are shown as mean ± SD. * P < .05, ** P < .01, **** P < .0001, by Student’s t test.

Figure 5.

ALPK1 mediates NF-κB-dependent responses to F. nucleatum infection. (a) Western blot analysis of ALPK1 in HCT116 cells treated with F. nucleatum, E. coli or PBS control. (b-c) Quantitative RT-PCR (b) and Western blot analysis (c) of ALPK1 were performed in HCT116 cells transfected with two siRNAs targeting ALPK1 or control siRNAs. (d) HCT116 cells transfected with the indicated siRNAs were co-cultured with F. nucleatum or PBS control. Western blot analysis of ALPK1, phosphor-IκBα, p65, phospho-p65 and ICAM1 were performed. (e) The distribution of NF-κB p65 in the indicated cells was measured by immunofluorescence assay. Scale bar, 20 μm. (f-g) HCT116 cells transfected with the indicated siRNAs were co-cultured with F. nucleatum or PBS control. Adhesion assay (f) and migration assay (g) were performed. The migrated cells were observed at 18 hours. Scale bar, 100 μm. (h-i) ALPK1-depleted HCT116 cells transfected with the indicated plasmids were applied for Western blot analysis (h) and adhesion assay (i). Scale bar, 100 μm. Data are shown as mean ± SD. *** P < .001, **** P < .0001, by Student’s t test.

Figure 6.

Overabundance of F.nucleatum correlates with high expressions of ALPK1 and ICAM1 in CRC patients and indicates poor clinical outcome. (a) C57BL/6 mice pretreated with antibiotics were administrated with F. nucleatum or PBS control everyday by gavage and sacrificed after the treatment at 15 days. (b) Quantitative RT-PCR analysis of the relative abundance of F. nucleatum in colon tissues from mice with the indicated treatment (n = 6). Data are normalized to Universal Eubacteria 16S. (c) Quantitative RT-PCR analysis of ICAM1 mRNA expression in colon tissues from the indicated mice (n = 6). (d) Representative IHC images of ICAM1 protein expression in colon tissues from the indicated mice. (e-f) Quantitative RT-PCR analysis of ALPK1 (e) and ICAM1 (f) in Cohort 1 which was divided into two groups depends on the relative abundance of F. nucleatum. (g) The correlation between the relative mRNA levels of ALPK1 and ICAM1 in Cohort 1. (h) The relative mRNA levels of ICAM1 among CRC patients with different lymph nodes metastasis stages in Cohort 1. (i-j) IHC analysis of ALPK1 (i) and ICAM1 (j) in Tissue Array Cohort 2. Representative photographs of IHC staining in CRC tissues with or without liver metastasis were shown (left), and the relative expression was quantified by H-score (right). Scale bar, 50 μm. (k-l) Kaplan-Meier survival curve of ALPK1 (k) and ICAM1 (l) expression of CRC patients in Tissue Array Cohort 2. (m) The relative abundance of F. nucleatum in the feces of CRC patients with positive lymph nodes metastasis (N1+N2, n = 32) or without metastasis (N0, n = 40) in Cohort 3. Data are shown as mean ± SD. * P < .05, ** P < .01, *** P < .001, **** P < .0001, Mann-Whitney test (b), Student’s t test (c, e, f, h, i, j, m), Linear Regression (g), Log-rank test (k, l).

Figure 7.

Schematic diagram of F. nucleatum-induced ALPK1/NF-κB/ICAM1 axis regulating the metastasis of CRC. The high abundance of F. nucleatum in gut activates NF-κB signaling of CRC cells through a newly identified pattern recognition receptor ALPK1, thereby promotes the expression of ICAM1, which is critical for F. nucleatum-induced CRC cell-endothelial cell adhesion, extravasation and metastasis.