A GNAQ/11-driven zebrafish cancer model identifies MITF and YAP as key determinants for uveal melanoma

Abstract

Uveal melanoma (UM) is a cancer of eye melanocytes that has a particularly poor prognosis and a dearth of treatment options. In contrast to cutaneous melanoma (CM), which is driven by oncogenic mutation of BRAF or NRAS, UM is driven by oncogenic mutations in GNAQ or GNA11 (>80% of patients). We developed a zebrafish model of UM by expressing GNAQ[superscript Q209L] in the melanocytes of p53-/- zebrafish. In this model, oncogenic GNAQ activates biological programs involved in pigmentation and tumorigenesis. The melanocyte lineage transcription factor MITF plays a well-defined role in CM tumorigenesis, wherein MITF expression is absolutely necessary for tumorigenesis and regulates proliferative versus invasive tumor phenotypes. In contrast, here we elucidated a novel tumor suppressor function for MITF in UM. In the context of oncogenic GNAQ, loss of mitfa accelerates tumorigenesis and reduces tumor reliance on p53 mutation.

Expression of oncogenic GNAQ can also overcome mitfa-/- tumor inhibition in BRAF-driven CM. Moreover, mitfa's tumor suppressive role is generalizable to UM and mitfa-/- accelerates tumorigenesis driven by oncogenic GNA11 or CYSLTR2, a GNAQ/11 upstream receptor. Interestingly, we show that GNAQ directly regulates pigment programs, as evidenced by hyperpigmentation patches that develop even in the absence of mitfa. Furthermore, the pigmentation and tumorigenesis phenotypes are decoupled, suggesting differential regulation of these programs downstream of oncogenic GNAQ. The observation that mitfa has different roles in CM vs UM led us to investigate the different signaling mechanisms of BRAF- vs GNAQ-driven melanoma. Oncogenic GNAQ activates the downstream pathways YAP and PLC[beta], and there is conflicting evidence implicating the relative importance of these two signaling axes in driving UM tumorigenesis. Here, we outline a central role for YAP for in in vivo UM tumorigenesis.

Constitutively active YAP[superscript S127A;S381A] drives rapid UM tumorigenesis, furthermore, YAP staining was ubiquitous across UM tumors. Finally, we show here for the first time that PLCB4D630Y was sufficient to drive tumors at long latency. Together, these findings clarify the genetic pathways that drive UM formation.