Fibroblasts potentiate melanoma cells in vitro invasiveness induced by UV-irradiated keratinocytes.

Melanoma represents a malignant disease with steadily increasing incidence. UV-irradiation is a recognized key factor in melanoma initiation. Therefore, the efficient prevention of UV tissue damage bears a critical potential for melanoma prevention. In this study, we tested the effect of UV irradiation of normal keratinocytes and their consequent interaction with normal and cancer-associated fibroblasts isolated from melanoma, respectively. Using this model of UV influenced microenvironment, we measured melanoma cell migration in 3-D collagen gels. These interactions were studied using DNA microarray technology, immunofluorescence staining, single cell electrophoresis assay, viability (dead/life) cell detection methods, and migration analysis. We observed that three 10 mJ/cm2 fractions at equal intervals over 72 h applied on keratinocytes lead to a 50% increase (p < 0.05) in in vitro invasion of melanoma cells. The introduction cancer-associated fibroblasts to such model further significantly stimulated melanoma cells in vitro invasiveness to a higher extent than normal fibroblasts. A panel of candidate gene products responsible for facilitation of melanoma cells invasion was defined with emphasis on IL-6, IL-8, and CXCL-1. In conclusion, this study demonstrates a synergistic effect between cancer microenvironment and UV irradiation in melanoma invasiveness under in vitro condition.

Simultaneous blocking of IL-6 and IL-8 is sufficient to fully inhibit CAF-induced human melanoma cell invasiveness.

Tumour microenvironment plays a critical role in cell invasion and metastasis. To investigate the role of cancer-associated fibroblasts (CAFs) in melanoma cell invasiveness, we used 3D spheroid invasion assay. The effect of conditioned media from normal fibroblasts and CAFs cultivated alone or co-cultivated with melanoma cells on BLM or A2058 melanoma spheroid invasion was analysed. We found that conditioned media from CAFs and CAFs co-cultured with melanoma cells, especially, promote invasion and migration, without significant effect on melanoma cell proliferation. We further analysed the expression of pro-invasive cytokines IL-8 and IL-6 in media and found that melanoma cells are dominant producers of IL-8 and fibroblasts are dominant producers of IL-6 in 2D monocultures, while co-cultivation of CAFs with melanoma cells induces production/secretion of IL-6 and IL-8 into the media. The analyses of IL-6 levels in 3D cultures and human melanoma samples, however, revealed that at least in some cases IL-6 is also produced directly by melanoma cells. Analysis of the role of IL-6 and IL-8 in CAF-induced melanoma invasion, using neutralising antibodies, revealed that simultaneous blocking of IL-6 and IL-8 is sufficient to fully inhibit CAF-induced human melanoma cell invasiveness. In summary, these experiments indicate the important role of CAFs and IL-8 and IL-6 cytokines in melanoma cell invasiveness.

Reduced WNT5A signaling in melanoma cells favors an amoeboid mode of invasion.

Tumor cells invade and spread via either a mesenchymal or an amoeboid mode of migration. Amoeboid tumor cells have a rounded morphology and pronounced RhoA activity. Here, we investigate how WNT5A signaling, a tumor promotor in melanoma, relates to Rho GTPase activity and amoeboid migration. We compared melanoma cells with low (HTB63 cells) and high (WM852 cells) WNT5A expression. HTB63 cells exhibited an amoeboid morphology and had higher RhoA activity but lower invasiveness than WM852 cells in a three-dimensional (3D) collagen matrix. We next explored the relationships between WNT5A, morphology, and invasive behavior. WNT5A knockdown impaired Rho GTPase Cdc42 activity, resulting in reduced invasion of amoeboid and mesenchymal melanoma cells. Interestingly, knockdown of WNT5A or inhibition of its secretion in WM852 cells expressing wild-type BRAF also led to increased RhoA activity via decreased RND3 expression, resulting in predominantly amoeboid morphology. In contrast, such treatments had the opposite effects on RND3 expression and RhoA activity in HTB63 cells expressing the active BRAFV600 mutation. However, treatment of HTB63 cells with a BRAF inhibitor made them respond to WNT5A knockdown in a similar manner as WM852 cells expressing wild-type BRAF. We next found that dual targeting of WNT5A and RhoA more effectively reduced melanoma cell invasion than targeting either protein individually. Taken together, our results suggest that low WNT5A signaling in melanoma cells promotes a rounded amoeboid type of invasion, which quite likely serves as a compensatory response to decreased WNT5A/Cdc42-driven invasion. This phenomenon partially explains the enduring melanoma cell invasion observed after impaired WNT5A signaling and has therapeutic implications. Our results suggest that dual targeting of WNT5A and RhoA signaling is a more effective strategy for controlling the invasion of BRAF wild-type and BRAFV600 mutated melanomas treated with a BRAF inhibitor than targeting either of the proteins individually.