Modeling of human uveal melanoma in zebrafish xenograft embryos.

PURPOSE: Uveal melanoma (UM) is fatal in up to 50% of patients because of liver metastases that are refractory to therapies currently available. While murine xenograft models for human uveal melanoma are available, they have limited utility for screening large compound libraries in drug discovery studies. Therefore, new robust preclinical models are needed that can efficiently evaluate drug efficacy for treatment of this malignancy. METHODS: Uveal melanoma cell lines generated from primary tumors (92.1, Mel270) and metastases (OMM2.3, OMM2.5, OMM1) were injected into the yolk of 2-day-old zebrafish embryos. After 6 days, proliferation and active migration was quantified via automated confocal image analysis. To determine the suitability of this xenotransplantation model for drug testing, drugs with three different activities (dasatinib, quisinostat, and MLN-4924) were added to the water of uveal melanoma-engrafted embryos. RESULTS: All tested UM cell lines proliferated and migrated in the embryos; significant differences could be discerned between cell lines: Cells derived from metastases showed more migration and proliferation than cells derived from the primary tumors, and provided preclinical models for drug testing. Addition of the Src-inhibitor dasatinib in the water of engrafted embryos reduced proliferation and migration of high Src-expressing 92.1 cells, but did not affect low Src-expressing metastatic OMM2.3 cells. Two experimental anticancer drugs, quisinostat (a histone deacetylase inhibitor) and MLN-4924 (neddylation pathway inhibitor), blocked migration and proliferation of 92.1 and OMM2.3. CONCLUSIONS: We established a zebrafish xenograft model of human uveal melanoma with demonstrated applicability for screening large libraries of compounds in drug discovery studies.