Polycomb Repressive Complex 2 (PRC2) is a group of proteins that catalyzes Histone 3 Lysine 27 (H3K27) tri-methylation (H3K27me3). The core subunits EZH2, EED, SUZ12 and RbAp4/7 are required for catalytic activity. Alterations in PRC2 have been identified in numerous human cancers. The histone methyltransferase EZH2 is overexpressed in breast and prostate cancer, and activating Y641 mutations have been identified in lymphoma. Inactivating EZH2 mutations occur in myeloid malignancies, and the histone variant H3.3K27M mutation that inhibits EZH2, resulting in a global loss of H3K27me3, is present in pediatric glioblastoma. This suggests that there are cell-type specific functions PRC2 in tumorigenesis. In melanoma, Y641 mutations and EZH2 amplification are found in 1% and 6% of cases, respectively. However, the exact role of PRC2 in melanoma has yet to be determined. To understand the function of EZH2 in melanoma, we used our existing zebrafish melanoma model. Zebrafish melanocytes that express oncogenic BRAF(V600E) and concurrently harbor a p53 loss of function develop melanoma. We ectopically expressed EZH2, EZH2-Y641F/N, H3.3 and H3.3K27M in zebrafish BRAF(V600E);p53-/- melanocytes and compared the rate of tumor formation. Surprisingly, H3.3K27M accelerated melanoma onset, but overexpression of wild-type H3.3, EZH2 or EZH2 mutants Y641F/N did not. This data suggests that loss of EZH2 methyltransferase activity promotes melanoma initiation in collaboration with BRAF(V600E), but gain of function mutants in EZH2 do not. To our knowledge these are the first in vivo data demonstrating the effect of PRC2 deregulation in an animal model of melanoma. We hypothesize that PRC2 activity loss alters the expression of bivalent genes, which are genes marked by activating (H3K4me3) and repressive (H3K27me3) histone marks. The expression of bivalent genes is tightly regulated because the genes are critically involved in the maintenance of proper differentiation. To investigate the role of EZH2 loss of function, we will use human melanoma cells and our zebrafish melanoma model. We will characterize the proliferative, epigenetic and gene expression changes that result using an inducible system in human melanoma cells. Next, we will integrate our epigenetic and gene expression results with human melanoma data to generate a set of candidate bivalent genes that are altered upon H3.3K27M expression. These genes will be tested using our melanoma model to identify downstream effectors of H3.3K27M oncogenicity. These studies will elucidate the function of PRC2 in melanoma, which has not been investigated.
The incidence of melanoma, the most lethal form of skin cancer, is increasing more rapidly than any other solid tumor, and metastatic disease is incurable. We used our zebrafish model that is genetically similar to human disease to determine that PRC2 loss of function accelerates melanoma formation, which to our knowledge is the first in vivo data demonstrating that PRC2 deregulation contributes to melanoma formation. In this proposal we will use a combination of zebrafish models and human cells to characterize the effects of PRC2 loss of function in melanoma in a comprehensive manner.