Identifying novel mechanisms of melanoma metastasis through embryonic melanoblast transcriptome analysis Open Access
Downloadable ContentDownload PDF
The enhanced ability of melanoma cells to metastasize is reminiscent of the innate propensity of melanoblasts to migrate to distant sites during embryonic development - from the neural crest to their eventual niche in the skin. Once transformed, melanoma cells mimic migratory and growth capabilities similar to those of embryonic melanoblasts. This putative relationship between tumorigenesis and developmental processes was first suggested by Rudolf Virchow more than 150 years ago. While this theory is largely unproven, there are mechanistic links between the processes regulating development and malignancy. Here we investigate this age-old puzzle using a mouse model with melanocyte-specific Green Fluorescent Protein (GFP) expression for the capture and sequencing of embryonic melanoblasts. We have, for the first time, isolated and sequenced the transcriptomes of murine embryonic melanoblasts at several key representative developmental stages. To uncover the overall classes of gene expression and to identify and characterize genesets whose expression is common and equally important to melanomagenic and developmental processes, a heat-map of the top 1000 most variable developmental genes was generated, and then shortlisted based on compared levels of expression in human and mouse metastatic melanomas, and on the relationship with melanoma patient survival data. By integrating bioinformatics and functional data we have devised a novel cross-species Multi-dimensional Embryonic-onco-Genomics Analyses (MEGA) approach. We show that late stage melanomas reactivate genes -which we refer to as metafetal genes- that are used during embryonic development to achieve a more aggressive metastatic phenotype. One such identified gene, encoding KDEL (Lys-Asp-Glu-Leu) endoplasmic reticulum (ER) protein retention receptor 3 (KDELR3), was highly expressed in advanced mouse melanomas relative to normal skin or to benign nevi. We confirmed this in human melanoma by showing that this metafetal gene was expressed at high levels in metastatic patient samples relative to benign lesions, and also predicted patient survival. Next, we determined the consequences of RNAi-based knockdown on experimental metastasis in mouse models. We validated the functional significance in human and mouse melanomas by showing that this KDELR3 plays a key role in melanoma metastasis through adaptation to chronic ER stress for survival by modulating the unfolded protein response (UPR) and autophagic components, which can be therapeutically targeted. Our study identifies key hardwired pathways associated with melanocyte development that can be co-opted by opportunistic metastatic melanoma cells. This approach offers a novel perspective on melanoma therapeutics and intervention and offers both mechanistic as well as prognostic insights into our understanding of this fatal disease.