Mucinous ovarian carcinoma (MOC) is a unique subtype of ovarian cancer with an uncertain etiology, including whether it genuinely arises at the ovary or is metastatic disease from other organs. In addition, the molecular drivers of invasive progression, high-grade and metastatic disease are poorly defined. We perform genetic analysis of MOC across all histological grades, including benign and borderline mucinous ovarian tumors, and compare these to tumors from other potential extra-ovarian sites of origin. Here we show that MOC is distinct from tumors from other sites and supports a progressive model of evolution from borderline precursors to high-grade invasive MOC. Key drivers of progression identified are TP53 mutation and copy number aberrations, including a notable amplicon on 9p13. High copy number aberration burden is associated with worse prognosis in MOC. Our data conclusively demonstrate that MOC arise from benign and borderline precursors at the ovary and are not extra-ovarian metastases.

MYB oncogene upregulation is associated with estrogen receptor (ER)-positive breast cancer, but disease requirements for MYB function in vivo have not been explored. In this study, we provide evidence of a critical requirement for MYB functions in models of human and murine breast cancer. In human breast cancer, we found that MYB expression was critical for tumor cell growth both in vitro and in vivo in xenograft settings. In transgenic knockout mice, tissue-specific deletion of the murine MYB gene caused a transient defect in mammary gland development that was reflected in delayed ductal branching and defective apical bud formation. In mouse mammary tumor virus (MMTV)-NEU mice where tumors are initiated by activation of HER2, MYB deletion was sufficient to abolish tumor formation. In the more aggressive MMTV-PyMT model system, MYB deletion delayed tumorigenesis significantly. Together, the findings in these transgenic knockout models implied that MYB was critical during an early window in mammary development when it was essential for tumor initiation, even though MYB loss did not exert a lasting impact upon normal mammary function. Two important MYB-target genes that promote cell survival, BCL2 and GRP78/BIP, were each elevated compared with nontransformed mammary epithelial cells, thereby promoting survival as confirmed in colony formation assays in vitro. Taken together, our findings establish a role for MYB at the hub of ER-and HER2-dependent pathways in mammary carcinogenesis. Cancer Res; 71(22); 7029-37. Ó2011 AACR.

Low-grade serous ovarian carcinoma (LGSOC) is associated with a poor response to existing chemotherapy, highlighting the need to perform comprehensive genomic analysis and identify new therapeutic vulnerabilities. The data presented here represent the largest genetic study of LGSOCs to date (n = 71), analysing 127 candidate genes derived from whole exome sequencing cohorts to generate mutation and copy-number variation data. Additionally, immunohistochemistry was performed on our LGSOC cohort assessing oestrogen receptor, progesterone receptor, TP53, and CDKN2A status. Targeted sequencing identified 47% of cases with mutations in key RAS/RAF pathway genes (KRAS, BRAF, and NRAS), as well as mutations in putative novel driver genes including USP9X (27%), MACF1 (11%), ARID1A (9%), NF2 (4%), DOT1L (6%), and ASH1L (4%). Immunohistochemistry evaluation revealed frequent oestrogen/progesterone receptor positivity (85%), along with CDKN2A protein loss (10%) and CDKN2A protein overexpression (6%), which were linked to shorter disease outcomes. Indeed, 90% of LGSOC samples harboured at least one potentially actionable alteration, which in 19/71 (27%) cases were predictive of clinical benefit from a standard treatment, either in another cancer's indication or in LGSOC specifically. In addition, we validated ubiquitin-specific protease 9X (USP9X), which is a chromosome X-linked substrate-specific deubiquitinase and tumour suppressor, as a relevant therapeutic target for LGSOC. Our comprehensive genomic study highlighted that there is an addiction to a limited number of unique 'driver' aberrations that could be translated into improved therapeutic paths.