Assessment of the oncogenic activity of melanoma-associated mutations in a human-tissue model of melanocytic neoplasia

The development of malignant melanoma is associated with several specific genetic changes. These changes often lead to activation of the Ras oncogene and two of its effector pathways, Raf-MEK-ERK and PI3K-Akt; inhibition of the p16INK4a/CDK4/Rb and p14ARF/HDM2/p53 tumor suppressor cascades; and telomerase activation. Oncogenic RAS mutations activate both Raf and PI3K cascades. The Raf-MEK-ERK pathway may also be activated via BRAF mutations, while PI3K-Akt signaling may be induced through PTEN loss or AKT3 amplification. Alterations at the CDKN2A locus, which encodes both p16 INK4a and p14ARF, may ablate one or both of these proteins. Rb bypass can also occur via activating CDK4 mutations, while p53 disruption can result from dominant-negative TP53 mutations. However, the extent to which these genetic changes can induce human melanocytic neoplasia is unknown.;To characterize the oncogenic potency of melanoma-associated genetic elements, we engineered human melanocytes to express genes of interest and incorporated the engineered cells into regenerated human skin tissue on immunodeficient mice. Co-expression of oncogenic RasG12V and hTERT (the catalytic subunit of human telomerase) induced junctional melanocytic hyperplasia, whereas co-expression of RasG12V, hTERT, constitutively active CDK4 R24C (which inactivates Rb), and dominant-negative p53R248W induced invasive melanocytic neoplasia with cardinal clinical, histologic, and immunophenotypic features of human melanoma.;We next used the genetically defined melanoma model to assess the functional impact of several genetic pathways. Co-expression of RasG12V and hTERT with either CDK4R24C or p53R248W produced invasive melanocytic tumors, suggesting that disruption of either Rb or p53 is oncogenic in this context. hTERT expression was dispensable for the development of early-stage melanocytic hyperplasia but required for progression to invasive neoplasia. Finally, expression of active B-RafV600E in place of RasG12V only led to junctional melanocytic hyperplasia, while active PI3K or Akt3 could replace RasG12V to induce invasive melanocytic neoplasia, suggesting that the PI3K-Akt pathway plays a central role in melanoma development. Thus, we have characterized a minimal set of genetic alterations sufficient to generate human melanocytic neoplasia and shown that genetically engineered human tissue provides a means of testing the oncogenic activity of genetic alterations identified in human cancer specimens.