Extracellular Vesicles From Pathological Microenvironment Induce Endothelial Cell Transformation and Abnormal Angiogenesis via Modulation of TRPV4 Channels

Abstract

The soluble and mechanical microenvironment surrounding endothelial cells influences and instructs them to form new blood vessels. The cells in the pathological tumor microenvironment release extracellular vesicles (EVs) for paracrine signaling. EVs have been shown to induce angiogenesis by communicating with endothelial cells, but the underlying molecular mechanisms are not well known. We have recently shown that the mechanosensitive ion channel transient receptor vanilloid 4 (TRPV4) expression and activity is significantly reduced in tumor endothelial cells (TEC), and that activation of TRPV4 normalized the tumor vasculature and improved cancer therapy. However, whether and how the tumor microenvironment downregulates TRPV4 and transforms the normal endothelial cell phenotype remains unknown. To explore this, we exposed normal human endothelial cells (hNEC) to human lung tumor cell conditioned media (TCM) and measured phenotypic changes and angiogenesis. We found that treatment with TCM transformed hNEC to a TEC-like phenotype (hTEC) as evidenced by increased expression of tumor endothelial cell marker 8 (TEM8) and exhibition of abnormal angiogenesis on 2D-Matrigels compared to normal hNEC. Mechanistically, expression and activity of TRPV4 was decreased in hTEC. Further, when pre-treated with exosome inhibitor GW4869, TCM failed to induce hNEC transformation to hTEC. Finally, addition of purified EVs from TCM induced transformation of hNEC to hTEC as evidenced by abnormal angiogenesis in vitro. Taken together, our results suggest that the pathological (tumor) microenvironment transforms normal endothelial cells into a tumor endothelial cell-like phenotype through EVs via the downregulation of TRPV4.

Keywords: TRPV4; angiogenesis; endothelial cells; extracellular vesicles; tumor.

FIGURE 1

Tumor cell conditioned media (TCM) induces transformation of hNEC to tumor endothelial-like (hTEC) phenotype. (A) qPCR analysis showing relative TEM8 gene expression in untreated hNEC and TCM-treated hNEC (hTEC). Note, TEM8 mRNA expression is increased significantly in hTEC (^∗p ≤ 0.05). Gene expression was normalized to β-actin and calculated as relative expression to hNEC. (B) Phase contrast images (4×) showing the angiogenic behavior of hNEC and hTEC when plated on 2D Matrigel at high density (100,000/well) at 6 and 24 h. Scale bar = 200 μm. (C) Quantitative analysis showing a significant decrease (^****p ≤ 0.0001) in tube length in hTEC compared to hNEC. The results shown are a mean ± SEM from three independent experiments.

FIGURE 2

TCM exposure downregulates functional expression of TRPV4 channels in hTEC. (A) Western blot analysis showing decreased expression of TRPV4 protein in hTEC when compared to hNEC. (B) Quantitative analysis of western blots revealed significant (^∗p ≤ 0.05) reduction of TRPV4 expression in hTEC. (C) Average traces showing calcium influx in response to the TRPV4 agonist, GSK1016790A (100 nM), in Fluo-4 loaded hNEC and hTEC. Arrow indicates the time of stimulation with TRPV4 agonist. (D) Quantitative analysis of calcium influx showed significant ^****p ≤ 0.0001 reduction in TRPV4-mediated calcium influx in hTEC compared to hNEC (F/F0 = ratio of normalized fluorescent intensity relative to time 0). The results shown are a mean ± SEM from three independent experiments.

FIGURE 3

hNEC transformation is correlated with the reduction in perinuclear VEGFR2 which is normalized by Rho kinase inhibitor Y27632. (A) Representative immunofluorescence images (60×) showing VEGFR2 localization in hNEC and hTEC. Cells were stained for DAPI (blue/nuclei) and total VEGFR2 (red) after exposure to TCM in 9 consecutive passages (hTEC). Scale bar = 10 μm. Note a strong perinuclear localization of VEGFR2 (arrows) in hNEC which was reduced in hTEC. (B) The graph represents the quantitative analysis of the percentage of cells with perinuclear localization of VEGFR2 showing a significant (^****p ≤ 0.0001) decrease in perinuclear VEGFR2 staining of hTEC. (C) Phase contrast micrographs (4×) showing the normalizing effects of the Rho kinase inhibitor, Y-27632 (10 μM; Y27) on hTEC angiogenic behavior, when plated on Matrigel at high density (100,000/well). Representative images were taken 24 h post plating, showing the formation of stable tubes upon inhibition of Rho kinase in hTEC. Scale bar = 200 μm. (D) Quantitative analysis of tube length showing a significant decrease (^****p ≤ 0.0001) in hTEC tube length compared with hNEC which was significantly increased (^****p ≤ 0.0001) after treatment with the Rho kinase inhibitor. The results shown are a mean ± SEM from three independent experiments.

FIGURE 4

Exosome inhibitor reduces formation of EVs and abolishes TCM-induced EC transformation. (A) Nanoparticle tracking analysis (NTA) showing reduction in the number of EVs from tumor cells that had been pre-treated with the exosome inhibitor GW4869 (GW). (B) Phase contrast micrographs (4×) showing normalized tube formation in hTEC exposed to GW4869 and plated on 2D Matrigels for 24 h compared to untreated hTEC and hNEC. Sale bar = 200 μm. (C) Quantitative analysis showing a significant increase (^****p ≤ 0.0001) in tube length between hTEC and hTEC + GW (GW4869) cells. Note a significant decrease in tube formation between hNEC and hTEC (^****p ≤ 0.0001), and no statistical difference (NS) between hNEC and hTEC + GW. The results shown are a mean ± SEM from three independent experiments.

FIGURE 5

Tumor-derived extracellular vesicles (EVs) induce hNEC transformation to hTEC as evidenced by abnormal angiogenesis. NTA analysis (A) and TEM images (B) showing 50–200 nm rounded structures confirming the isolation EVs from TCM. Scale bar = 50 nm. (C) Phase contrast micrographs (4×) showing tube formation of control and EV-treated cells plated on 2D Matrigel at 24 h. Scale bar = 200 μm. (D) Quantitative analysis showing a significant decrease (^****p ≤ 0.0001) in tube length by EC (hTEC) treated with tumor-derived EVs for 48 h. The results shown are a mean ± SEM from three independent experiments.