Chemotactic factors underlying tumor infiltration by immunocompetent cells in human colorectal cancer

Abstract

Colorectal cancer (CRC) is a common digestive tract malignancy and a major cause of cancer mortality. Several studies have convincingly shown that CRC infiltration by immunocompetent cells and, in particular, cytotoxic CD8+ T cells (CTLs), IFN-γ-producing T-helper 1 cells (Th1), Foxp3+ regulatory T cells (Tregs), and CD16+ MPO+ neutrophils, is significantly associated with prolonged patient survival. However, the chemotactic factors driving these cell populations into the tumor site, their cellular sources and their microenvironmental triggers remain to be elucidated. During my PhD training I have investigated the chemokine/chemokine receptor network promoting CRC infiltration by immune cells associated to favorable prognosis. In particular, I addressed: 1. The expression of immune cell markers and their correlation with chemokine expression in primary CRC tissues; 2. The identification of chemokine receptors relevant for CRC infiltration by beneficial immune cells; 3. The chemokine sources in CRC; 4. The microenvironmental stimuli triggering chemokine production in CRC tissues; 5. The effects of chemokine production on immune cell recruitment into CRC. The expression of a panel of genes encoding 39 chemokines and 7 markers specific for defined immune cell populations was assessed by quantitative PCR array in 62 samples of freshly excised primary CRC and autologous healthy colonic tissue. Correlations between expression of chemokine genes and immune cell markers were then evaluated. Furthermore, chemokine receptor profiles were analysed by flow cytometry on cell suspensions obtained upon digestion of clinical specimens or on corresponding cell populations from autologous peripheral blood. Based on chemokine receptor expression on tumor infiltrating cells and correlations between expression of chemokines and immune cell markers, I could identify for each immune cell subset a putative “chemokine signature”: 1) CCL3, CCL5, CCL8 CXCL9, CXCL10 and CXCL12, associated with recruitment of cytotoxic CTLs; 2) CCL5, CCL22, CXCL9, and CXCL12 correlating with infiltration by Th1; 3) CCL22 and CXCL12 potentially attracting Tregs; 4) CXCL2 and CXCL5 promoting chemotaxis of CD16+ MPO+ neutrophils. I have further investigated potential chemokine sources and stimuli leading to chemokine release within CRC tissues. I found that CRC cells purified from primary tumor specimens express many of the genes encoding identified immune cell recruiting chemokines, including CCL3, CCL5, CXCL2, CXCL5, CXCL9 and CXCL10. In vitro experiments showed that chemokine production by CRC cells is triggered upon their exposure to microbial stimuli, such as Toll-like receptor agonists, or CRC-associated bacteria, including Fusobacterium nucleatum, Bacteroides Fragilis, Bacteroides vulgatus, and Escherichia Coli, thus suggesting that components of the gut flora may critically influence chemokine production in CRC tissues. This was indeed confirmed by “in vivo” experiments showing that chemokine gene expression in xenografts, generated upon injection of human CRC cells in immunodeficient NSG mice, appeared to be related to the presence of commensal bacteria. In particular, chemokine gene expression levels in intracecal xenografts, were found to be ≥10 fold higher as compared to those of subcutaneous xenografts, and they were significantly reduced upon antibiotic treatment of tumor bearing mice. Most importantly, a correlation between extent of immune cell infiltration and bacterial load was also observed in human CRC samples. Indeed, CRC samples characterized by high expression of chemokine and immune cell markers, displayed significantly higher bacterial loads, as assessed by analysis of bacterial 16S ribosomal RNA, as compared to samples showing low chemokine expression and immune cell infiltration. In addition, a significant correlation between bacterial load and expression of the Th1 marker IRF1, CCL3 and CCL5, was also detected. Our in vitro and in vivo results cumulatively suggest that bacteria-induced chemokine production by tumor cells may lead to tumor infiltration by beneficial immune cells. Consistent with this hypothesis, in preliminary “in vitro” experiments, I found that supernatants of bacteria-stimulated CRC cells promote chemotaxis of CTLs and Th1 cells to a higher extent than untreated tumor cells. Additional “in vivo” studies are clearly warranted. In particular, I plan to evaluate intratumoral recruitment of CRC-derived CTLs and Th1 cells upon adoptively transfer into intracecal xenografts-bearing mice. Bacterial species or strains mostly contributing to high chemokine expression and immune cell infiltration in human CRC samples also remain to be identified. Microbiome analysis of CRC samples characterized by high or low immune cell infiltration might be envisaged in future studies. The results of the present work together with the proposed additional studies will contribute to the understanding of the interplay occurring between gut flora and immune system in CRC, and may pave the way towards innovative treatments aimed at modifying the gut flora in order to promote CRC infiltration by beneficial immune cell subsets.