FRACTALKINE-FRACTALKINE RECEPTOR AXIS IN HUMAN COLORECTAL CANCER

Although the presence of genetic and epigenetic abnormalities is important for tumor occurrence, the formation of an inflammatory microenvironment plays a pivotal role in tumor progression. Even if the link between cancer and inflammation has been widely accepted in the last decade, no inflammation-associated markers or potential targets have yet been translated into clinical practice. Human colorectal cancer (CRC), the second cause of cancer-related death in Western countries, represents a paradigm of the cancer-related inflammation. The aim of this study is to provide an overview of the inflammatory mediators expressed in tumor microenvironment of CRC, in order to select and characterize new candidates for clinical or therapeutic strategies. In this study, the RNA profile of a large panel of inflammatory genes, in particular cytokines, chemokines and receptors, has been analysed in eight surgical tumor samples and corresponding normal mucosa, from CRC patients. Several chemokines and receptors were found to be strongly up regulated in tumor tissues. The expression of selected genes, such as CCL3, CCL4, CXCL8 and CX3CL1, was further investigated and confirmed in a larger number of samples. Further studies focus on a specific chemokine-receptor pair: CX3CL1 and CX3CR1. CX3CL1 is a transmembrane protein acting both as an adhesion molecule as well as a chemokine when cleaved by specific protease. It has a single receptor, CX3CR1, which has been primarily studied on immune cells. A number of reports indicate that the CX3CL1-CX3CR1 axis is involved in the pathogenesis of different chronic inflammatory conditions, but its role in cancer has not been elucidated. By immunohistochemistry, we analysed the expression of CX3CL1 and CX3CR1 in a series of 144 paraffin-embedded CRC tissues. We found that CX3CL1 and CX3CR1 were already faintly expressed by healthy epithelial cells, and their expression progressively increased in pre-cancerous lesions and in tumor tissues, with a heterogeneous staining. Based on the median score value of CX3CL1 and CX3CR1 staining, we divided our case list in high and low expressing tumors. Correlating these results with patients’ clinical and pathological features, we found that patients with high expression of CX3CL1 and CX3CR1 showed a trend to better prognosis. Of great interest, the concomitant high expression of both CX3CL1 and CX3CR1 (positive-axis tumors) resulted in significantly increased disease-specific survival and disease-free survival. Given the role of CX3CL1 in the recruitment of immune cells at sites of inflammation, we investigated whether high expression of CX3CL1 correlated with an increased amount of tumor infiltrating lymphocytes (CD3+ cells, known to be protective in CRC) or macrophages (CD68+ cells). To our surprise, CX3CL1 expression by tumor cells did not correlate either with the density of CD3+ nor CD68+ cells at tumor invasive front, indicating that the CX3CL1-associated favourable prognosis is not mediated via the recruitment of these immune cells. Since CX3CL1 is able to act as an adhesion molecule for CX3CR1+ cells, we hypothesized that the concomitant expression of CX3CL1 and CX3CR1 by tumor epithelial cells could function as an adhesive loop, increasing the adherence among cells and preventing tumor spreading. This hypothesis is supported by our findings that CRC metastatic tumor cells in lymph nodes and liver showed a reduced expression of CX3CL1 and CX3CR1 in comparison to the corresponding primary tumors. More interesting, patients with positive-axis tumor showed a reduced risk of developing metachronous metastasis. To have insights into the functional interaction of the two molecules, we ectopically overexpressed CX3CL1-cherry fusion protein and CX3CR1-GFP fusion protein in two different human colorectal cancer cell lines. In vitro studies demonstrated that the expression of CX3CR1 and CX3CL1 strongly increased tumor cell adhesion but not cell proliferation. Using a multicellular tumor spheroid (MTS) model obtained by mixing CX3CL1-cherry+ cells with CX3CR1-GFP+ cells, we observed by immunofluorescence the presence of several spots of colocalization, confirmed by FRET analysis, suggesting that tumor cells indeed interact in a CX3CL1-CX3CR1-mediated mechanism. To gain support on our hypothesis regarding the putative protective role of CX3CL1-CX3CR1 in preventing tumor metastasis, in vivo experiments have been undertaken. Of note, the presence or absence of the CX3CL1-CX3CR1 axis did not influence tumor growth (subcutaneous) in nude mice. A syngenic model of spleen-liver metastasis is in progress. Taken together, these data indicate that the presence of CX3CL1-CX3CR1 axis is a good prognostic marker in CRC patients. The lack of this axis is associated with an increased risk of tumor spreading and metastasis. The mechanistic explanation of these results still needs to be elucidated. Our observation on the role of CX3CL1-CX3CR1 axis in CRC suggests a new biological significance of the involvement of chemokines in tumors. While the presence of chemokines and their receptors in cancer has been usually associated with tumor progression, we demonstrate in this study that the expression of the intact CX3CL1-CX3CR1 loop by CRC epithelial cells is associated with lower degree of tumor recurrence and metastasis. These results suggest a different point of view in the paradigm of chemokines and cancer-related inflammation.