Low grade serous ovarian tumours are a rare and under-characterised histological subtype of epithelial ovarian tumours, with little known of the molecular drivers and facilitators of tumorigenesis beyond classic oncogenic RAS/RAF mutations. With a move towards targeted therapies due to the chemoresistant nature of this subtype, it is pertinent to more fully characterise the genetic events driving this tumour type, some of which may influence response to therapy and/or development of drug resistance. We performed genome-wide high-resolution genomic copy number analysis (Affymetrix SNP6.0) and mutation hotspot screening (KRAS, BRAF, NRAS, HRAS, ERBB2 and TP53) to compare a large cohort of ovarian serous borderline tumours (SBTs, n = 57) with low grade serous carcinomas (LGSCs, n = 19). Whole exome sequencing was performed for 13 SBTs, nine LGSCs and one mixed low/high grade carcinoma. Copy number aberrations were detected in 61% (35/57) of SBTs, compared to 100% (19/19) of LGSCs. Oncogenic RAS/RAF/ERBB2 mutations were detected in 82.5% (47/57) of SBTs compared to 63% (12/19) of LGSCs, with NRAS mutations detected only in LGSC. Some copy number aberrations appeared to be enriched in LGSC, most significantly loss of 9p and homozygous deletions of the CDKN2A/2B locus. Exome sequencing identified BRAF, KRAS, NRAS, USP9X and EIF1AX as the most frequently mutated genes. We have identified markers of progression from borderline to LGSC and novel drivers of LGSC. USP9X and EIF1AX have both been linked to regulation of mTOR, suggesting that mTOR inhibitors may be a key companion treatment for targeted therapy trials of MEK and RAF inhibitors.

No abstract

BackgroundMucinous ovarian tumors are an unusual group of rare neoplasms with an apparently clear progression from benign to borderline to carcinoma, yet with a controversial cell of origin in the ovarian surface epithelium. They are thought to be molecularly distinct from other ovarian tumors but there have been no exome-level sequencing studies performed to date.MethodsTo understand the genetic etiology of mucinous ovarian tumors and assess the presence of novel therapeutic targets or pathways, we undertook exome sequencing of 24 tumors encompassing benign (5), borderline (8) and carcinoma (11) histologies and also assessed a validation cohort of 58 tumors for specific gene regions including exons 4–9 of TP53.ResultsThe predominant mutational signature was of C>T transitions in a NpCpG context, indicative of deamination of methyl-cytosines. As well as mutations in known drivers (KRAS, BRAF and CDKN2A), we identified a high percentage of carcinomas with TP53 mutations (52 %), and recurrent mutations in RNF43, ELF3, GNAS, ERBB3 and KLF5.ConclusionsThe diversity of mutational targets suggests multiple routes to tumorigenesis in this heterogeneous group of tumors that is generally distinct from other ovarian subtypes.Electronic supplementary materialThe online version of this article (doi:10.1186/s13073-015-0210-y) contains supplementary material, which is available to authorized users.

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.

Endometrial secretions in the uterine cavity contain mediators important for endometrial receptivity and embryo implantation. Unbiased analysis of uterine fluid from a receptive versus nonreceptive time of the menstrual cycle and in fertile and infertile women will provide new insights into uterine receptivity. We hypothesized that proteomic analysis of human uterine lavages would identify proteins important for the establishment of pregnancy in humans. Lavages collected from fertile (n = 7) and infertile (n = 8) women during the midsecretory (MS) phase, and from fertile women during the midproliferative (MP) (n = 7) phase, were assessed using 2D-differential in gel electrophoresis (2D-DiGE) over a pI 4-7 range. Statistical analysis revealed 7 spots that were significantly decreased in the MP compared to the MS phase, while 18 spots showed differential expression between fertile and infertile women. A number of proteins were identified by mass spectrometry, including antithrombin III and alpha-2-macroglobulin, whose production was confirmed in endometrial epithelium. Their staining pattern suggests roles during embryo implantation. Assessment of the human endometrial secretome has identified differences in the protein content of uterine fluid with respect to receptivity and fertility.

This study aimed to identify proteins differentially expressed in the human endometrium between the proliferative and secretory phases of normal menstrual cycles by 2D differential in-gel electrophoresis (DIGE). A total of 196 out of 1017 spots were differentially expressed (p < 0.05). Mass spectrometry identified 76 proteins representing 41 different gene products. Immunohistochemistry confirmed the observed changes in 3 representative proteins (Rho-GDIalpha, CLIC1, PGRMC1). Biological pathway analysis identified the Jnk and EGF signaling pathways as key regulators of protein expression in the midsecretory phase of endometrial proteome.

Endometriosis is a chronic disorder affecting approximately 10% of women in whom endometrial tissue forms painful lesions outside the uterus. It has a major impact on their physical, mental and social well-being but has no known cure, and there is no nonsurgical means of diagnosis. We have used a proteomic approach to identify proteins with altered abundance in the eutopic endometrium of endometriosis patients in the midsecretory phase of the menstrual cycle. 2D-differential in gel electrophoresis (DIGE) and mass spectrometry identified 20 proteins that were present at different levels in endometriosis patients (p < 0.05), many of which have not previously been associated with endometriosis. Protein abundance changes did not correlate well with published gene array data, emphasizing the extensive post-translational modification that occurs in this tissue. Abundance or localization changes in endometrial tissue were validated by immunohistochemistry and Western blotting for three proteins, vimentin (VIM), peroxiredoxin 6 (PRDX6), and ribonuclease/angiogenin inhibitor 1 (RNH1), while observed changes could not be confirmed for coronin 1A (CORO1A) or transgelin (TAGLN2). In addition, multiple charge and size isoforms were observed for PDRX6 and vimentin (VIM), and an additional PDRX6 isoform was observed in endometriosis patients that was below the level of detection in healthy women. Biological pathway analysis identified that cytoskeletal remodeling via keratin intermediate filaments, processing of the cystic fibrosis transmembrane receptor (CFTR), the glucocorticoid receptor subunit alpha (GCR), and heat shock factor 1 (HSF1) were all significantly over-represented features in endometriosis patients. This study highlights the highly dynamic nature of endometrial tissue and suggests that considerable post-translational modification of proteins is a key factor in the pathology of endometriosis.

No abstract