Identification of markers for human endometrial carcinoma stem cells

Abstract

Cancer stem cells (CSCs) have recently been identified in many human solid cancers. It was hypothesised that CSCs are responsible for epithelial neoplasia associated with endometrial carcinoma (EC), the most common gynaecological malignancy in women. The aim of this study was to demonstrate that a rare population of EC cells possess CSC properties, and to commence identifying markers for these cells. Firstly, evidence for a subset of cells with CSC properties in human EC was required. Functional assays revealed that a small population (0.24%) of freshly isolated EC and endometrial hyperplasia cells exhibited the in vitro CSC characteristics of clonogenicty and self renewal. Further assays indicated that rare cells within EC were able to initiate tumourigenic growth, recapitulating the original tumour phenotype when transplanted subrenally in NOD/SCID mice. Importantly, this project discovered that not all cells within EC possessed the same clonogenic, tumour initiating, and self renewal properties, conforming to the CSC hypothesis in all grades of Type I EC, Type II EC and its precursor lesion, endometrial hyperplasia. However, these studies were retrospective and a cell surface marker or combination of markers for the isolation of endometrial cancer stem cells (ECSCs) is required so that they can be isolated, characterised and their role in the development of the disease studied. To identify potential ECSC markers, advantage was taken of the heterogeneity of EC cells. The hypothesis was that ECSCs would express different markers from their progenitor and more differentiated daughter cancer cells. A panel of 26 potential adult stem cell antibody supernatants was investigated. Since none of the antibodies showed a pattern of reactivity suggestive of CSCs i.e. strong reaction with a small number of EC epithelial cells, a strategy was devised to rank the antibodies into a priority list. This priority list was based on the percentage of cells expressing the marker, the marker’s robustness, and consistency of expression between samples. Based on this priority list, 2 antibodies were tested using functional assays to indicate if they selected for ECSCs. The CD133-1 antibody (W6B3) was identified as the top potential ECSC marker. Interestingly, the number 3 ranked antibody was the CD133-2 antibody (293-C3). Reports were appearing that indicated there were differences in the expression of these two CD133 epitopes, and thus both were concurrently investigated to determine if either were potential ECSC markers. Neither antibody enriched for epithelial EC cells with in vitro CSC properties. Similar percentages of clonogenic cells (0-0.5%) were observed in the CD133+ and CD133- (CD133-1 or CD133-2) sorted populations, indicating no enrichment of colony forming units (CFUs). Further, CFUs from both CD133+ and CD133- subpopulations were equally able to undergo self renewing divisions in vitro, with no observable difference between CD133-1 and CD133-2. This data indicated that CD133 did not enrich for ECSCs. This study lays the groundwork for future studies to explore further markers from the priority list that will enable the prospective isolation of ECSC required to confirm their existence and role in the development of human EC. Together with the identification of normal human endometrial epithelial stem cell markers, the development and progression of EC can be investigated, allowing the identification of potential drug targets selective for CSC, but sparing normal endometrial stem cells. Such treatments will be particularly useful for early stage EC and hyperplasia candidates where the uterus may be conserved, and for late stage cases where hysterectomy is not curative and current treatments target the bulk tumor cells rather than CSCs.