Two novel genetically engineered mouse models for uveal melanoma with liver metastasis

Uveal melanoma (UM) is the most common malignant tumor of the eye. It arises from melanocytes in the choroid, iris, or ciliary body. Approximately 50% of patients will develop one or more metastases and thereafter survive an average of 6–12 months, due to a lack of effective systemic therapy. The most common site of metastasis is the liver (93%), followed by the lungs (30%). All UMs activate Gαq/₁₁ signaling, mostly through gain-of-function mutations in either GNAQ or GNA11, which encode heterotrimeric G protein alpha subunits. These mutations are also common in other non-epithelium melanocytic neoplasms such as blue nevi and leptomeningeal melanocytomas. Gαq/₁₁ activation is an early event in disease progression. In contrast, inactivating mutations in the gene encoding a deubiquitination protein, BAP1, are enriched in liver metastases and predict a poor patient outcome. To model the in vivo effects of oncogenic GNAQ with Bap1 loss, we generated a transgenic mouse model which expresses oncogenic GNAQQ²⁰⁹ᴸ from the ubiquitously expressed Rosa26 locus in all melanocytes beginning during embryogenesis, in combination with either somatic and/or germline deletion of Bap1. All GNAQQ²⁰⁹ᴸ expressing mice recapitulated the Gαq/11 human disease spectrum, developing UM, leptomeningeal melanocytoma, and blue nevus-like dermal hyper-pigmentation. 40% of the mice with somatic heterozygous Bap1 loss and 20% of the mice with germline heterozygous Bap1 loss also developed a single, heavily pigmented liver metastasis by 5 months of age. No liver lesions were found in any of the other mice in the study. Hence, we have generated the first mouse model that replicates human UM progression. These two mouse models can now be used to dissect the molecular mechanisms driving UM liver metastasis and to develop therapeutic treatments for advanced disease.Medicine, Faculty ofMedical Genetics, Department ofGraduat