Glioblastoma is the most aggressive primary brain tumor, and, despite intensive studies, remains one of the most fatal malignancy in adult humans. Among multiple onco-promoters produced by glioblastoma cells, erythropoietin was found. However, the presence/function of Epo alternatively spliced variants in cancer remains unexplored. Here, we investigated the expression and role of Epo-variants in glioblastoma, and the therapeutic potential of their targeting through a novel monoclonal antibody (mAb).

Transcripts and protein levels of Epo-variants in a cohort of human brain tumors were evaluated by RT-PCR, ELISA, and immunohistochemistry. Monoclonal antibodies targeting Epo-Vs were prepared and functionally selected by assaying proliferation, migration, stemness, and angiogenesis in glioblastoma patient-derived cells. Antibody affinity for Epo/Epo-variant was determined by SPR. In vivo toxicity and therapeutic efficacy of the lead antibody were evaluated in GBM mouse models.

We found a significant overexpression of Epo-variant transcripts in tissues and cells from GBM patients. After functional selection of newly-produced antibodies, we identified AND-C4 as the lead one for its potent anti-tumoral properties, absence of anti-erythropoietic effects and of toxicity on human brain cells. AND-C4 exhibited high affinity for the Epo-variant EV-3. We demonstrated that EV-3 was efficiently produced and secreted by glioblastoma cells, particularly by stem cells. EV-3 exerted tumorigenic, angiogenic and immunomodulatory properties, and AND-C4 was effective in antagonizing all these actions. In vivo studies in rodent glioblastoma models revealed that AND-C4 selectively bound to tumor tissue and exhibited significant efficacy on tumor growth and animal survival.

This study represents the first evidence on the presence, origin and pro-tumoral activity of EV-3 in human glioblastoma. Moreover, in vitro and in vivo results revealed AND-C4 as novel and promising anti-glioblastoma immunotherapeutic.