Function and evolution of the piRNA pathway in the amniote gonad and human ovarian cancer.

Abstract

The Piwi-interacting RNA (piRNA) pathway is a RNA silencing pathway which represses the expression of gene and transposable elements (TE) in the gonads via the binding of piRNAs (26-30nt) to their complimentary RNA targets and by influencing the epigenetic makeup of chromatin via interacting with other proteins (e.g. HP1). piRNAs interact with PIWI (P-element induced wimpy testis) proteins and other components such as Maelstrom (Mael) for TE silencing. In addition, Piwi-like (Piwil) genes and Mael are important for germline stem cell (GSC) production from fly to mouse. The expression of these genes was reported exclusively in mammalian testis despite the presence of piRNAs in mouse testis and ovary. Although the pathway is essential for oogenesis in fly, fish and Xenopus, an important role in the mammalian ovary has been in doubt, as female knockout mice (Piwil1, 2 and Mael) genes are fertile. In addition Piwil genes in particular have undergone lineage specific changes leading to up to 4 Piwil genes in most eutherian mammals. Work presented in this thesis investigates the expression of piRNA pathway genes in the amniote gonad. This confirmed the robust expression of Piwil genes and Mael in the mammalian testis. Importantly, specific expression of these genes in oocyte and growing follicles was detected in mammals. The extraordinary conservation of piRNA pathway gene expression in germ cells and ovarian somatic cells from fly to human suggests an important role in mammalian gonadal development. A comprehensive bioinformatics analysis of Piwil genes provided new aspects towards understanding the evolutionary trajectory of Piwi like genes in vertebrates. For example the correction of Xenopus piwil3 as piwil1 ortholog clearly showing that Piwil3 evolved exclusively in eutherian mammals. Finally, based on the expression of piRNA pathway genes in the ovary, we hypothesised that PIWIL genes and MAEL may play a part in the origin and progression of epithelial ovarian cancer (EOC). To test this hypothesis, expression of these genes was investigated in postovulatory tissues i.e. the corpus luteum (CL) and inclusion cysts. Preliminary data show that piRNA pathway genes are not expressed in the CL but transcripts are detected in the epithelial cells of inclusion cysts. This raises the possibility that piRNA pathway genes may be involved in the cancerous transformation of epithelial cells. To test if the piRNA pathway plays a part in EOC progression, and is related to the activity of TEs, the expression of PIWIL genes, MAEL and L1 (one of the most abundant TEs in human) was investigated in different types of EOC. Significant upregulation of these genes was found in malignant EOC when compared to benign ovarian tumours. This upregulation might be a result of increased L1 activity in EOC, or may due to the stem cell like characteristic of malignant EOC. Analyses of the PIWIL1 transcripts from a malignant EOC show that most of the PIWIL1 transcripts contain premature stop codons. Therefore, although PIWIL1 is overexpressed in malignant EOC, the function of PIWIL1 is likely to be compromised in these tumours. To understand the effects of PIWIL1 and MAEL overexpression on cancer cell invasiveness, PIWIL1 and MAEL were transiently overexpressed in ovarian cancer SKOV3 cells. Overexpression of these genes decreases cell invasiveness, suggesting a repressive role in EOC progression. Hypomethylation of L1 and chromosome instability was found in ovarian cancers. To understand the genome stability in EOC, primary ovarian cancer cells were established from patient ascites of different stages of ovarian cancers. FISH analyses showed that 20%-40% of the cells are aneuploid. Thus, this is a good model for understanding aneuploidy in EOC development. Our research provides a better understanding of this ancient yet conserved piRNA pathway in mammalian gonads and ovarian cancers.