Uveal Melanoma-Derived Extracellular Vesicles Display Transforming Potential and Carry Protein Cargo Involved in Metastatic Niche Preparation

Abstract

Extracellular vesicles (EVs) carry molecules derived from donor cells and are able to alter the properties of recipient cells. They are important players during the genesis and progression of tumors. Uveal melanoma (UM) is the most common primary intraocular tumor in adults and is associated with a high rate of metastasis, primarily to the liver. However, the mechanisms underlying this process are poorly understood. In the present study, we analyzed the oncogenic potential of UM-derived EVs and their protein signature. We isolated and characterized EVs from five UM cell lines and from normal choroidal melanocytes (NCMs). BRCA1-deficient fibroblasts (Fibro-BKO) were exposed to the EVs and analyzed for their growth in vitro and their reprograming potential in vivo following inoculation into NOD-SCID mice. Mass spectrometry of proteins from UM-EVs and NCM-EVs was performed to determine a protein signature that could elucidate potential key players in UM progression. In-depth analyses showed the presence of exosomal markers, and proteins involved in cell-cell and focal adhesion, endocytosis, and PI3K-Akt signaling pathway. Notably, we observed high expression levels of HSP90, HSP70 and integrin V in UM-EVs. Our data bring new evidence on the involvement of UM-EVs in cancer progression and metastasis.

Keywords: Uveal melanoma; extracellular vesicles; liquid biopsy; liver metastasis; mass spectrometry.

Figure 1

Characterization of extracellular vesicles (EVs) derived from normal choroidal melanocytes (NCMs) and Uveal Melanoma (UM) cells. (A,B) Nanoparticle tracing analysis NTA of EVs derived MEL270, metastatic UM cells (OMM2.5) and NCMs. (B) NTA data showing concentrations of EVs from different cell sources. Data are expressed as mean ± SD (n = 3). (C) Representative micrographs of immunoGold-TEM on MP46-EVs (Ci–Cii) 92.1-EVs (Ciii) and NCM-EVs (Civ) labelled with a cocktail of antibodies against CD81 (Ci), TSG101 (Cii) and CD63 (Ciii) (red arrows). Scale bars 200 nm. (D) Proteins isolated from EVs derived from different cell sources were analyzed by Western blot for the expression of specific exosome markers.

Figure 2

Isolated EVs were efficiently internalized by target cells. EVs from UM cells and NCMs were labeled with PKH67 dye and purified using Optiprep dentity gradient and ultracentrifugation. PKH67-labeled EVs were added to Fibro-BKO cells (A–D) and IHH (F–I). EV uptake (red arrows) was monitored under confocal microscopy 6 h later. (E,J) Cells were exposed to PKH67 solution processed as was the case for labeled EVs. Scale bars 20 μm.

Figure 3

UM-EVs promote proliferation, migration and invasion of BRCA1-deficient fibroblast (Fibro-BKO) cells. (A–C) Fibro-BKO cells were cultured for 3 weeks in the presence of culture medium without EVs (No-EVs), NCM-EVs (from PPWc and PPxG eye donors) or UM-EVs. (A,B) Cells were analyzed for their growth potential by measuring population doubling capability at every passage. Data in inserts represent cumulative population doublings at the end of the treatment periods (A). Column graphs represent pooled data from three EV-free medium (No-EVs) samples, two NCM-EVs preparations and five UM-EVs preparations. Data are mean ± SD. p values ˂ 0.05 (*) (B). (C–F) Fibro-BKO cells were cultured for 12 h (C,D; cell migration) or 24 h (E,F; cell invasion) in the presence of culture medium without EVs (PBS), NCM-EVs or UM-EVs. Data are mean  ±  SD. (D) p value = 0.0086 (**), p value = 0.0051 (#). (F) p value = 0.0072 (**), p value < 0.0001 (****), MEL270 UM-EVs/OMM2.5 UM-EVs, p value = 0.0033 (**) (n = 3). Scale bar:100 µm, magnification 100×.

Figure 4

In vivo tumorigenicity assay of Fibro-BKO cells treated with UM-EVs (A)Exposed cells were injected subcutaneously into NOD/SCID mice that were monitored for 4 weeks for tumors growth. At euthanasia, developing tumors were excised and their sizes were measured. (B) Formalin-fixed paraffin-embedded tumors were processed for H&E staining, and immunolabeled with anti-Ki67, anti-Vimentin and anti-MelanA antibodies. Scale bar: 10 µm. Red arrowheads pointed to mitotic figures, and black arrow pointed to a melanophage. Positive controls are from choroidal melanoma specimens.

Figure 5

UM-EVs and NCM-EVs carried different sets of proteins. Venn diagram analyses. (A) The majority of proteins isolated from EVs derived from UM cells and NCMs were shared with data published in Vesiclepedia database. (B) NCM-EVs and primary UM-EVs shared 708 proteins, while 1 and 1433 proteins were exclusively present in NCM-EVs and primary UM-EVs, respectively. (C) Based on their protein cargo, primary UM-EVs clustered differently from NCM-EVs. A total of 20 to 265 proteins were exclusively shared between NCM-EVs and primary UM-EVs, while 287 to 1074 proteins were exclusively shared between EVs isolated from the different primary UM lines.

Figure 6

Primary UM-EVs are enriched in proteins involved in the regulation of tumor growth, homeostasis and metastasis organotropism. (A) Volcano plot representation of 308 proteins significantly and differentially expressed between primary UM-EVs and NCM-EVs. (B) Heatmap chart representing the 308 differentially expressed proteins. Note that primary UM-EV contents clustered differently from that of NCM-EVs. The full list of proteins is shown in Table S2. (C) Heatmap chart depicting the relative expression levels of proteins linked to tumorigenesis, cancer homeostasis, and metastasis organotropism. (D) Immunoblot validation of the relative expression levels of some proteins that emerged from MS data mining (see C). Note: HSP90 shown (Figure S7 (E)) was probed on a different membrane. In B,C, the key color represents Log(2) of protein quantitative ration where blue and red refer to downexpressed and overexpressed proteins in UM-EVs, respectively.

Figure 7

Gene ontology GO classification of proteomic data for differentially expressed proteins in primary UM-EVs and NCM-EVs. The most enriched categories in (A) biological process, (B) molecular function, and (C) cellular component. Left panel—UM-EVs, right panels—NCM-EVs.

Figure 8

Primary UM-EV protein cargo clustered differently from that of metastatic UM-EVs. (A,B) Venn diagram analyses. Primary and metastatic UM-EVs shared 676 proteins, while 868 and 86 proteins were exclusively present in primary MEL270 UM-EVs and metastatic OMM2.5 UM-EVs, respectively (A). Based on their protein cargo, primary UM-EVs clustered differently from metastatic UM-EVs. 47 to 89 proteins were exclusively shared between primary UM-EVs and metastatic UM-EVs, while 508 to 704 proteins were exclusively shared between EVs isolated from the different primary UM lines (B). (C) Volcano plot representation of 262 proteins significantly and differentially expressed between primary and metastatic UM-EVs (MEL270 vs. OMM2.5). (D) Heatmap chart representing the 262 differentially expressed proteins. Note that primary MEL270 UM-EV contents clustered differently from that of metastatic OMM2.5 UM-EVs. The full list of proteins is shown in Table S3. (E) Heatmap chart depicting the relative expression levels of proteins linked to metastasis organotropism and metastasis regulation. In (D,E), the key color represents Log(2) of protein quantitative ration where blue and red refer to downexpressed and overexpressed proteins, respectively.

Figure 9

GO classification of proteomic data for differentially expressed proteins in primary MEL270 UM-EVs and metastatic OMM2.5 UM-EVs. The most enriched categories in (A) biological process, (B) molecular function, and (C) cellular component. Left panel—primary UM-EVs, right panels—metastatic UM-EVs.