Dual effect of semaphorin 4d blockade in neuroendocrine tumor malignizationfrom vessels to macrophages

Resumen

Tumor progression is a complex conundrum of events that involve not only tumor cells, but also their surrounding microenvironment. Accounting to the natural dependency of tumors on angiogenesis, its therapeutic targeting remains a valid stroma-directed strategy in the fight against cancer. However, standard antiangiogenics fail to produce enduring beneficial effects due to the appearance of resistance, often as a consequence of intratumor hypoxia triggered by vessel trimming. In the case of neuroendocrine tumors (NETs), which are characterized by their low aggressiveness, high heterogeneity and vessel content, new therapeutic approaches are being explored to overcome such hurdles, where conventional therapies stumble. In this context, blockade of semaphorin 4D (Sema4D), a proangiogenic molecule with homeostatic roles in the immune system, by a monoclonal antibody (anti-Sema4D) has proved beneficial antitumor effects in the RIP1-Tag2 preclinical mouse model of pancreatic NET (PanNET). Unfortunately, the decrease in tumor burden and increase in survival of anti-Sema4D treated mice followed an increase in local invasiveness and metastasis. Contrary to the thoroughly described mechanisms governing malignization after antiangiogenic therapies, no intratumoral hypoxia was detected after Sema4D blockade. In this doctoral thesis, aiming to decipher this novel form of resistance by which anti-Sema4D treatment acts as a double-edged sword in PanNETs, the two systems involved in Sema4D signaling were studied: the vascular and the immune system. We first described a beneficial antiangiogenic effect, characterized by structural changes in tumor vessels mediated by a pericyte-endothelial cell crosstalk. Incidentally, we found that the aggressive phenotype involved the recruitment of Sema4D-positive tumor-associated macrophages (TAMs) to the tumor ecosystem, which, after becoming activated by anti-Sema4D treatment, triggered tumor cell migration and invasion. Mechanistically, functional characterization of Sema4D-positive TAMs’ secretome revealed cytokine CXCL12 to be one of the molecules involved in the proinvasive program, suggesting the implication of CXCL12/CXCR4 signaling. Comprehensive clinical validation further shed light on the implication of both macrophage-derived Sema4D and CXCR4 in the malignization steps of tumor development in PanNET patients, which undoubtedly unleashes a new range of approaches merging the immunotherapy and the antiangiogenic fields in their shared fight against cancer.