Tumor-associated macrophages display differential protein cargo sorting in extracellular vesicles associated with poor survival in ovarian cancer.

Ovarian cancer (OC) progression and metastasis are promoted by ascites, which constitutes a central part of the tumor microenvironment (TME). In this fluid, tumor-associated macrophages (TAMs) represent a prominent immune cell type. In addition to tumor and other host cells such as TAMs, ascites is highly enriched in soluble factors as well as extracellular vesicles (EVs). How TAMs contribute to the EV compartment of the OC TME remains, however, underexplored. In this work peripheral blood monocytes from healthy donors were differentiated into monocyte-derived macrophages (MDMs) and polarized into classically activated (M1-like), alternatively activated (M2-like) and TAM-like (by ascites incubation). For all subtypes, serum-free conditioned medium was collected for 24 h and EVs were isolated and characterized by nano-flow cytometry (nFC), label-free mass spectrometry-based proteomics and electron microscopy, among others. Our results demonstrated distinct traits for EV release and cargo across the different macrophage subtypes. Specifically, TAM-like macrophages exhibited impaired release of small EVs and reduced frequency of tetraspanin-positive particles. These EV subpopulations displayed sizing profiles closer to M1-like than to M2-like samples. Also, the low EV release in TAM-like MDMs was accompanied by altered expression of biogenesis-related markers like flotillin-1 (FLOT1) and a decreased N-glycosylation of CD63 protein, which was validated in patient-derived samples. Remarkably, the EV-associated proteome of TAMs displayed significant enrichment in both pro- and anti-inflammatory molecules with clinical value. Markers significantly enriched in the ascites TAM-EV signature were mostly associated with poor prognosis, whereas M1-like EV-related markers (pro-inflammatory) were mostly associated with longer survival. Our results confirmed previous data for proteins like CD163 and MRC1 to be associated to TAM-EVs, while also describing novel candidates with diagnostic (i.e., COLEC12) and/or prognostic (i.e., MSR1) value in plasma. Taken together, our data support a unique secretory profile of TAMs in OC and provide new EV-associated biomarkers with translational impact. Our results pave the way for a better understanding of the mechanisms behind TAM-EV cargo loading and function, and how these cells participate in the TME landscape.