Kate Lawrenson

Loss of Scribble causes cell competition in mammalian cells.

In Drosophila, normal and transformed cells compete with each other for survival in a process called cell competition. However, it is not known whether comparable phenomena also occur in mammals. Scribble is a tumor suppressor protein in Drosophila and mammals. In this study we examine the interface between normal and Scribble-knockdown epithelial cells using Madin–Darby Canine Kidney (MDCK) cells expressing Scribble short hairpin RNA (shRNA) in a tetracycline-inducible manner. We observe that Scribble-knockdown cells undergo apoptosis and are apically extruded from the epithelium when surrounded by normal cells. Apoptosis does not occur when Scribble-knockdown cells are cultured alone, suggesting that the presence of surrounding normal cells induces the cell death. We also show that death of Scribble-knockdown cells occurs independently of apical extrusion. Finally, we demonstrate that apoptosis of Scribble-knockdown cells depends on activation of p38 mitogen-activated protein kinase (MAPK). This is the first demonstration that an oncogenic transformation within an epithelium induces cell competition in a mammalian cell culture system.

In vitro three-dimensional modelling of human ovarian surface epithelial cells.

Objectives:  Ninety percent of malignant ovarian cancers are epithelial and thought to arise from the ovarian surface epithelium (OSE). We hypothesized that biological characteristics of primary OSE cells would more closely resemble OSE in vivo if established as three‐dimensional (3D) cultures.

Interactions between normal and transformed epithelial cells: Their contributions to tumourigenesis

During the initial stages of carcinogenesis, neoplastic transformation occurs in single epithelial cells and the transformed cells proliferate while being surrounded by normal epithelia. In Drosophila, normal and transformed epithelial cells compete with each other for survival, a process called cell competition. However, it was not known whether comparable phenomena also occur in mammals. Recently, several reports have shown that the interaction between normal and transformed epithelial cells causes various phenomena in mammals. For example, with elaborate cell culture systems that express oncoproteins or knockdown tumour suppressor proteins in an inducible manner, certain types of transformed cells have been shown to be apically eliminated from normal epithelial layers in an apoptosis-dependent or -independent manner. During the process of apical extrusion, various signalling pathways are modulated in transformed cells located within the normal epithelium, indicating that the presence of surrounding normal epithelial cells affects the behaviour and fate of transformed cells. Recent studies in mice have also shown that normal and transformed cells can compete with each other for survival during several processes such as liver regeneration. In this review, we will introduce these recent publications on interactions between normal and transformed mammalian epithelial cells. Furthermore, we will discuss how these studies can potentially lead to identification of biomarkers for precancerous cells and to invention of novel types of cancer prevention and treatment.

Somatic Genetic Development in Epithelial Ovarian Cancer

The genetic and biological mechanisms that underlie the somatic development of epithelial ovarian cancers remain poorly understood. For many other tumour types, there is substantial evidence of an ordered series of stages in neoplastic development, driven by the accumulation of somatic genetic changes and the deregulation of specific biological pathways [1]. For ovarian cancer, there is still debate in the published literature about the tissue type from which epithelial ovarian cancers originate and the nature of the precursor lesions.

Abstract 5021: A 3D model of genetic transformation of normal ovarian epithelial cells identifies candidate genes associated with the initiation and development of epithelial ovarian cancer

Proceedings: AACR 101st Annual Meeting 2010‐‐ Apr 17‐21, 2010; Washington, DC Novel biomarkers associated with early-stage ovarian cancer (OC) development may represent susceptibility makers that initiate ovarian cancers or screening markers for early disease detection. Little is known about the biology underlying OC development and heterogeneity. It has been hyporthesis that OCs can arise from the ovarian surface epithelium (OSE), a monolayer of epithelial cells covering the surface of the ovary. We have shown that immortalised OSE cells (IOSE) bypass replicative senescence but show no other features of neoplastic transformation. Here, we established an in vitro three-dimensional (3D) genetic model of early-stage ovarian carcinogenesis. By first overexpressing the CMYC oncogene, and then expressing mutant alleles of the KRAS or BRAF oncogenes in IOSE cells, we were able to create a stepwise model of neoplastic transformation of OSE. Overexpression of CMYC in IOSE cells (IOSE-CMYC) induces a significant increase in anchorage-dependent and -independent growth, enhanced progression through the cell cycle and decreased apoptosis. Subsequent expression of mutant KRAS-G12V or BRAF-V600E alleles in IOSE-CMYC cells caused differential rates of anchorage-independent growth, invasiveness and proliferation showing that the different genetic elements mimics the phenotypic heterogeneity seen in primary ovarian tumors in vivo. 3D models of oncogene-expressing clones revealed many characteristics of malignant cells in vivo that could not be detected in 2D monolayer cultures. Gene expression microarray profiling of the different stages of this model identified several candidate tumour suppressor genes that are associated with neoplastic transformation. These included: THBS1, an inhibitor of angiogenesis in several cancer types; FEZ1, which is involved in cell growth and shows loss of expression in ∼40% of OCs; and RGS4, (a regulator of G-protein signaling) shown to promote breast cancer migration and invasion. Novel activated genes were also identified including PITX1, previously shown to be overexpressed in a murine model of ovarian granulosa cell tumours; and HIST1H4C, which is a overexpressed in therapy related myloid leukemias. We used the model to evaluate the functional role of genes that have recently been shown to increase susceptibility to ovarian cancer using genome wide association studies. One locus is 8q24, near the CMYC oncogene, highlighting its significance in OC development. Two other loci are BNC2 at 9q21 and TiPARP at 3q25. Both genes were down-regulated in response to CMYC over-expression and showed significant loss of function with further transformation after KRAS-G12V activation (p=7.8 × 10-7 and p=6.3 × 10-4, respectively). This suggests both behave as tumor suppressor genes. Note: This abstract was not presented at the AACR 101st Annual Meeting 2010 because the presenter was unable to attend. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 5021.

Abstract 1415: Stromal and epithelial interactions in three dimensional models of epithelial ovarian cancer

Proceedings: AACR 101st Annual Meeting 2010‐‐ Apr 17‐21, 2010; Washington, DC In vitro cell biology models of disease are a fundamental tool for the understanding of disease development in vivo. For cancer, two dimensional (2D) monolayer cell cultures have been used to establish in vitro models. However, it has become increasingly evident that these models fail to address important characteristics of tumours in vivo. The three dimensional (3D) architecture and the resulting interactions with the tumour microenvironment cannot be recreated in standard 2D cultures. We have shown that homotypic 3D models of normal ovarian surface epithelial (OSE) cells and epithelial ovarian cancers (EOCs) represent good models of tissues in vivo. However, heterotypic models including two or more cell types are likely to be even more relevant. Therefore we have created 3D models of ovarian stromal and epithelial cell interactions for EOC and the normal ovary. We established an immortalised ovarian fibroblast cell line from a normal ovary to represent the microenvironment in normal ovarian tissues. To recreate the stroma of a normal ovary in a postmenopausal woman we induced senescence in the normal ovarian fibroblasts (NOFs). Morphological and histocytochemical examination of spheroids with senescent ovarian fibroblasts (SOFs) showed increased numbers of mitosis and staining for Mib-1, suggesting that SOFs can promote epithelial cell proliferation. To model the stroma of a malignant tumour we used mesenchymal stem cells (MSC) which can differentiate into cancer associated fibroblasts (CAFs) after being recruited into the tumour stroma. We induced differentiation of MSC into a ‘CAF -like’ phenotype using conditioned medium from EOC cell line cultures. Differentiation of MSCs was evident one week after conditioning and was verified by staining for the CAF markers aSMA, FSP, Vimentin and FAP. Stromal and epithelial cells were labelled using eGFP and the far red fluorescent protein mKate2 respectively, to enable discrimination between both cell types. Heterotypic 3D cultures were set-up by co-culturing the stromal cells with transformed ovarian epithelial cells using polyHEMA coated tissue culture plastics. Fluorescent labelling has enabled us to study the 3D invasive properties and the influence of the changing microenvironment on the invasive ability of transformed epithelial cells. In conclusion, we have established 3D spheroid models of stromal and epithelial interactions representing both normal ovarian tissues and EOCs. These models will serve as valuable reagents in our understanding of stromal-epithelial interactions and the development of ovarian cancer. Note: This abstract was not presented at the AACR 101st Annual Meeting 2010 because the presenter was unable to attend. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 1415.

Abstract 3860: New ovarian cancer susceptibility loci identified

Genome-wide association analyses have identified an ovarian cancer susceptibility locus near BNC2 at 9p22.2 (OR 0.82, p 5.1 × 10 −19 , Nat Genet 41:996-1000, 2009), based on 1,772 and 2,354 Stage I cases and controls, 4,803 and 5,237 Stage 2 cases and controls, and extensive Stage 3 analysis in independent samples. Here, we report on Stage 3 analysis of nine additional regions with Stage 2 results at p −4 . Stage 1 samples were genotyped using the Illumina 610Quad and imputed at up to 2 million SNPs; Stage 2 samples were genotyped on an Illumina iSelect of 22,790 SNPs. Stage 3 samples from 4,335 cases (including 2,197 serous subtype) and 5,951 controls (14 studies) were genotyped using Illumina or Sequenom at up to 32 SNPs on chromosomes 1, 2, 3, 7, 8, 11, and 17. No evidence of heterogeneity by study site was observed. Logistic regression revealed four regions with ordinal p ≤ 10 −7 in all cases or serous subtype only. The most significant finding for all cases was for rs2072590 at 2q31.1 (all cases OR 1.15, p=8.1 × 10 −13 ; serous only OR 1.19 p=9.0 × 10 −13 ) which is located 3.8 kb 3-prime of HOXD3, a homeobox gene which also had an intronic SNP (rs2301301) with p=3.1 × 10 −5 (all cases). Second, we observed a locus at 8q24.21 highlighted by rs10088218 (all cases OR 0.84, p=1.3 × 10 −9 ; serous only OR 0.75 p=2.2 × 10 −14 ). This locus resides on the opposite side of MYC and 1.3 Mb from the 8q24 regions harboring loci for other cancers, and is independent of the other cancer-associated variants (r 2 −7 ; serous only OR 1.26, p=3.2 × 10 −8 ). Finally, at 17q21.32, rs9303542 was associated with risk particularly among serous cases (all cases OR 1.13, p=1.5 × 10 −6 ; serous only OR 1.16 p=4.3 × 10 −7 ); this SNP is intronic to SKAP1 (src kinase associated phosphoprotein 1). We evaluated the functional role of genes at these loci in ovarian cancer and normal ovarian epithelial cells and in an in vitro model of neoplastic initiation and progression from normal ovarian epithelial cells after CYMC and KRAS activation. Both BNC2 and TiPARP were down regulated in cancer versus normal cells (p=5 × 10 −9 and p=3.4 × 10 −5 , respectively) suggesting that loss of function is required for cancer development. Over-expressing CMYC (at 8q24) induced neoplastic transformation of normal ovarian cells. Both BNC2 and TiPARP were down-regulated following CMYC over-expression and showed significant loss of function with further transformation after mutant KRAS activation (p=7.8 × 10 −7 and p=6.3 × 10 −4 , respectively). These results emphasize the importance of genome-wide association studies and consortia for the identification of moderate-risk alleles in rare cancers. Importantly, our findings unravel the genetic components of this disease and suggest functionally-significant novel targets. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 3860.