We confirmed that EVs derived from melanoma, breast cancer and pancreatic cancer cell lines carry receptors for IFNγ, known as interferon-gamma receptor (IFNGR).
Additionally, we showed that IFNγ binds less effectively to EVs derived from melanoma cells that express lower levels of IFNGR (IFNGR⁻ EVs, generated by IFNGR1-knockdown), than to EVs derived from wild-type melanoma cells (IFNGR+ EVs). This confirms that melanoma EVs bind IFNγ, at least in part, via IFNGR.
To understand whether the IFNGR on EVs influences tumor growth and metastases formation, we treated mice bearing melanoma tumors with IFNGR+ EVs, IFNGR- EVs, or PBS (used as control). The tumor volume and metastatic spread to the draining lymph node and lungs did not significantly differ across treatment groups. However, in tumors from both EV-treated groups, we observed an increase in the percentage of immune cells, including CD8⁺ T cells and macrophages, compared to tumors from PBS-treated animals. These results suggest that EVs promote immune cell recruitment to the tumor microenvironment; however, the effect appears to be independent of IFNGR levels on the EVs.
We hypothesized that we could not distinguish the specific effects of IFNGR+ EVs vs IFNGR- EVs due to the continued production of IFNGR+ EVs by the melanoma cell line itself. To address that, we generated a melanoma cell line with reduced EV secretion (EV-inhibited cells) by knocking down RalB, a protein involved in exosome biogenesis.
We found that IFNγ was more cytotoxic to EV-inhibited cells than to wild-type melanoma cells, suggesting that EVs protect melanoma cells from IFNγ-mediated cytotoxicity in vitro. Importantly, we also observed that IFNGR⁺ EVs, but not IFNGR⁻ EVs, enhanced in vivo tumor growth, compared to PBS. However, the proportion of tumor and immune cells within the tumors was not significantly different between treatment groups.
To evaluate whether IFNGR⁺ EVs influence the effectiveness of ICI therapy, we treated tumor-bearing mice with PBS, IFNGR⁺ EVs alone, anti-PD-L1 antibody alone, or the combination of IFNGR⁺ EVs and anti-PD-L1. As expected, anti-PD-L1 treatment slightly reduced tumor growth, relatively to PBS. However, the combination therapy also decreased tumor volume, when comparing with PBS, suggesting that melanoma-derived IFNGR⁺ EVs do not compromise the efficacy of ICI therapy.
Altogether, our findings support the hypothesis that IFNGR on EVs is required to promote tumor growth and that EVs can protect tumor cells from the cytotoxic effects of IFNγ. However, further studies are needed to determine whether IFNGR on EVs influences ICI therapy outcomes and whether it could serve as a biomarker to stratify patients based on their likelihood of responding to treatment.