Search for ncRNAs that contribute to oncogene-induced senescence.

Tumour suppressor protein p53 contributes to various pathways involved in cell proliferation, cell death and normal cell homeostasis. It is not obvious which of these pathways is essential to block tumourigenesis; novel p53 targets and pathways may still contribute as key mechanisms to cancer inhibition.

We have identified previously unknown p53 targets in the class of long noncoding RNA through RNA expression profiling by qPCR as well as RNA sequencing. In vitro validation of these targets confirms p53-dependent induction, as well as direct binding of p53 to the promoter of these genes by chromatin immunoprecipitation. For the nuclear abundant long noncoding RNA NEAT1, we confirmed that transcription is induced in multiple cell types in conditions leading to p53 activation such as Nutlin treatment, chronic DNA stress and senescence.

NEAT1/Neat1 noncoding transcript, in human and mouse respectively, is the architectural component of nuclear bodies called paraspeckles, which sequester many paraspeckle proteins and RNAs. These structures can be formed in pathological conditions or disrupted in response to transcriptional stress. We found that NEAT1 induction in response to p53 activation also leads to increased formation of paraspeckles. We have also discovered that NEAT1 is required for cell survival in a breast cancer cell line. We noted that this effect can be partially attributed to a feedback mechanism in which NEAT1 decreases p53 protein levels. However, the components of this negative feedback remain to be identified.

NEAT1 is expressed at high levels in a majority of cell lines but in tissue samples the RNA levels are even higher. Moreover, we noted that these RNA levels are not correlated with the presence of paraspeckles. We have probed NEAT1 in a variety of human cancer samples and found high paraspeckle numbers in HER2 positive breast cancer and squamous cell carcinoma. We next investigated a panel of tissue samples containing benign human skin lesions of keratinocytic origin as well as squamous cell carcinoma. We also discovered that Neat1-paraspeckles are not detectable in normal mouse skin but are formed in all investigated pre-malignant tumours. However, upon progression into malignant carcinoma, Neat1-paraspeckles are only found in half of the samples.

Neat1 deficient mice show no apparent defects. However, when they are challenged by carcinogen induced mutations in skin keratinocytes, papilloma and subsequent carcinoma formation is substantially decreased compared to wild-type mice. A detailed study of mouse mammary gland development also showed that mice which are deficient for this long noncoding RNA have altered ductal morphogenesis and reduced lactation capacity due to decreased cell proliferation.

To identify relevant pathways downstream of NEAT1, we knocked it down in a breast carcinoma cell line and performed RNA sequencing in conditions with and without p53 stimulation. Transcriptome analysis revealed putative involvement of NEAT1 in the regulation of mitochondrial fission, STAT signalling and interferon type I response. Further studies will have to show how these responses are linked to NEAT1 or paraspeckles and which pathway is relevant for each condition or cell type.

Overall, this thesis characterises a previously annotated long noncoding RNA, NEAT1, in the p53 pathway. We showed that Neat1 is important for cell proliferation in normal breast development and for tumour initiation in skin keratinocytes. These findings indicate that Neat1 could be a therapeutic target to block cell growth.