Immunobiology of neutrophils in human colorectal cancer

Colorectal cancer (CRC) represents a major cause of cancer related death in different geographic areas. Tumor infiltration by a variety of immune cell types has consistently been observed to be associated with favorable prognosis (1). In particular, CRC infiltration by CD8+ T cells expressing memory and activation markers has repeatedly been reported to represent a favorable prognostic marker, although underlying molecular mechanisms and the antigen specificity of these cells are largely unclear. In sharp contrast, tumor infiltration by myeloid cells has classically been associated with poor prognosis in a variety of cancers of different histological origin (2). However, recent reports underline that granulocytes might participate to anti-tumor immune responses in defined cancer types (3-5). In previous studies, our group has observed that CRC infiltration by myeloid cells expressing CD16 and myeloperoxidase (MPO) is also associated with good prognosis (6;7). Based on this background, I have addressed the role of granulocyte infiltration in CRC immunobiology. By using a clinically annotated tissue microarray (TMA) including over 650 individual CRC, I have explored the prognostic significance of tumor infiltration by CD66b+ granulocytes. I found that CRC infiltration by CD66b+ granulocytes significantly correlates with favorable prognosis. Most importantly however, dense granulocyte infiltration significantly enhances the prognostic significance of CD8+ T cell infiltration in CRC. Taken together these data suggest that a cross-talk of potentially high clinical relevance takes place in CRC. Indeed, immunofluorescence staining of tissue sections from CRC showing evidence of infiltration by lymphocytes and granulocytes indicates that CD8+ and CD66b+ cells are frequently co-localized. Still unclear is whether neutrophils are able to modulate T-cell responsiveness to antigenic challenges (3;8). On the other hand activated T cells have previously been shown to enhance granulocyte survival and functions by the release of cytokines, including TNF-α, IFN-γ and GM-CSF (8). In additional studies I have extensively characterized phenotypic profiles of CRC infiltrating granulocytes, as compared to granulocytes from peripheral blood or infiltrating healthy mucosa adjacent to CRC. Functional studies performed by using tumor-derived and peripheral blood neutrophils indicate that these cells are able to efficiently co-stimulate CD8+ T cell activation induced by T-cell receptor (TCR) triggering, as witnessed by increased expression of CD69 activation marker, proliferation and IFN-γ release. Furthermore, an expansion of cells expressing a “central memory” phenotypic profile was also observed in cultures stimulated in the presence of autologous neutrophils. Furthermore, importantly, malignant transformation in the intestinal mucosa is associated with early translocation of microorganisms from the gut lumen (9). A variety of bacterial strains including Bacteroides fragilis and Fusobacterium nucleatum have been found to be highly represented in CRC (10). Capitalizing on these data, I have investigated the potential significance of the interaction with bacterial strains associated with CRC in the elicitation of the anti-tumor effects of neutrophils. I found that interaction with Fusobacterium nucleatum rapidly induces apoptosis in neutrophils and abrogates their co-stimulatory capacity. Taken together, my results contribute to the identification of CRC microenvironment as typically characterized by a “ménage à trois”, including cancerous cells, the immune system and gut colonizing microorganisms. The nature of the reciprocal interaction of these actors and its outcome are likely to decisively impact on CRC development and progressio