Contribution of tumor-associated neutrophils in the course of cutaneous squamous cell carcinoma : neutrophils and cancer

Non-melanoma skin carcinomas are the most frequent cancers in Human and their incidence is constantly increasing. Two main types exist: the cutaneous basal cell carcinoma (cBCC) and the cutaneous squamous cell carcinomas (cSCC). Risk factors include sun radiation and immunosuppression. These cancers are mainly treated with surgery and radiotherapy but they can reach an incurable stage. For this reason, novel therapeutic alternatives are needed. At present, immunotherapies constitute a revolution in the treatment of cancers. Its mechanism of action relies on the stimulation of the immune system of cancer patients, so that they develop efficient anti-tumoral immune responses. cSCC may benefit from this type of treatment as they generally develop in the context of an immunosuppression. Immune surveillance involves both immune cells and the tumor microenvironment, in particular the stroma. During the elimination phase, the anti-tumoral responses, mediated mainly by CD8+ T lymphocytes, are efficient. Then, there is an equilibrium phase in which the tumor is stable before the escape phase, when the tumor can evade immune surveillance and grow. Our research interest focused on neutrophils, a subset of myeloid cells that are very rapidly recruited to the sites of inflammation and inside tumors. A recent meta-analysis of 39 human malignancies showed that neutrophils are associated with the worst clinical outcome. Neutrophils harbor both anti- and pro-tumoral functions. This polarization seems to be dependent on type I interferon and TGF-β, respectively. It remains to establish the exact role played by neutrophils in cancer and specifically in skin carcinomas. The aim of our research was to further characterize the functions and the contribution of neutrophils to the development of cutaneous squamous cell carcinomas. We first used a chemically-induced skin carcinoma mouse model that recapitulates the different stages of skin carcinoma development in Human. In this model, we saw a massive infiltration of neutrophils at the precancerous and cancerous stages. We performed transcriptomic analysis of highly purified neutrophil populations from precancerous, cancerous lesions and from the surrounding skin controls. These data revealed a specific gene signature in neutrophils from lesions compared to surrounding skins. Differential gene expression analysis identified a pro-tumoral phenotype for neutrophils infiltrating lesions compared to skins. In a second approach, we studied the growth of a cSCC cell line grafted in the dermis of mice. Specific depletion of neutrophils significantly delayed tumor growth, thus indicating that neutrophils were pro-tumoral. Mechanisms of action included the production of ROS and NO that favor tumor growth and the immune suppression of anti-tumoral responses mediated by tumor-associated CD8+ T cells. In the tumor, neutrophils produced arginase 1 which catalyzes the degradation of arginine, thus inhibiting the proliferation of CD8+ T cells. In addition, we found that the tumor microenvironment induced PD-L1 expression at the cell surface of neutrophils and concomitantly, PD1 on CD8+ and CD4+ T cells. These results suggested that PD-L1/PD1 interaction triggers immune suppression and contributes to SCC progression. Indeed, a positive and significant correlation was observed between tumor size and frequencies of PD-L1-expressing neutrophils inside tumors. Collectively, these results suggest that it is relevant to assess immunotherapies that block PD-L1/PD1 interaction for the treatment for cSCC. These approaches could be combined with treatments that aim to block the recruitment or inhibit neutrophils. Moreover, it remains to evaluate whether the frequency of PD-L1-expressing neutrophils could constitute a good predictive marker of the response to anti-PD-L1 and anti-PD1 immunotherapies.