NONCOSENSE · Identifying long non-coding RNAs with a role in Oncogenic-Induced Senescence

Identifying long non-coding RNAs with a role in Oncogenic-Induced Senescence

1. A summary description of the project objectives: The objective 1 of this project was to identify functional long non-coding RNAs (lncRNAs) important for oncogenic-induced senescence (OIS). Once identified the candidate the second objective was to fully characterize the lncRNA in a descriptive and functional manner. And the third and more time-consuming objective was to identify the molecular mechanism underneath the function of the candidate lncRNA. 2. A description of the work performed since the beginning of the project: Objective 1: Identify functional lncRNAs important for oncogenic-induced senescence. We have follow 2 approaches to identify putative candidates: A custom designed microarray and RNA-sequencing. 1.A) Nimblegen lncRNA array. A custom array platform for lncRNAs designed in the host laboratory in collaboration with the Department of Molecular Medicine at Aarhus University Hospital has been used to compare human fibroblasts expressing a 4-OHT inducible allele of BRAF treated with EtOH (control) to 4-OHT (senescence). Analysis of these data led to the identification of several lncRNAs deregulated in senescence. 1.B) RNA-sequencing comparing human fibroblasts expressing a 4-OHT inducible allele of BRAF treated with EtOH (control) to 4-OHT (senescence). Experiments were performed in triplicates and send for sequencing to BGI (China). Jakob S. Pedersen and Morten M. Nielsen at Aarhus University performed bioinformatic analysis. 1.C) Validation of the candidate lncRNAs. Quantitative RT-PCR analyses have been used to validate the candidate lncRNAs that were identified by the methods described above. The functionality of the lncRNA in senescence has been assessed using siRNA and shRNA to remove the lncRNA. Also, overexpression experiments using lentiviral transduction has been performed. The senescence phenotype has been assessed by different methods such as proliferation analysis (crystal violet, MTT assays), Senescence Associated-β-galactosidase activity, p16INK4A staining, cell cycle analyses, senescence-associated heterochromatin foci (SAHF) detection, western blotting or qPCR. For the second part of the project I have optimized in vitro invasion assays to assess the invading capability of the cells and enzyme linked immunosorbent assays (ELISA) for the detection of interleukins in the media. 1.D) Expression analysis and clinical validation of the selected target. In collaboration with Anders Jacobsen (currently working at University of Copenhagen) we have analyzed melanoma cancer patient data available at The Cancer Genome Atlas (TCGA). Objective 2: For the descriptive analysis of the candidate lncRNA I have used standard molecular and cellular techniques to complete the description of the selected lncRNAs. 2.A) Validation of the lack of translational potential. I analyzed by qPCR polysome fractionation followed to determine that the lncRNA is not associated with polysomes and therefore not translated. 2.B) Determination of the exact length and sequence of the lncRNAs. I have used RT-qPCR using different primers to identify the isoforms and the length of the transcript. Besides, I have used public available data from ENCODE project to check isoforms and expression levels in different cell types. 2.C) Identification of the localization of the lncRNA. Cytoplasm-nucleus fractionation, followed by qRT-PCR was used to determine the localization of the transcript in the cell. Objective 3: Mechanistic analysis of the function of the lncRNAs selected. 3.A) Identification of the pathway in which the selected lncRNA is involved. We have performed RNA sequencing from control and siMIR31HG depleted cells in order to identify mRNAs that are deregulated upon MIR31HG depletion. The samples were sent to BGI and China and analyze by Jakob S. Pedersen and Morten M. Nielsen at Aarhus University. 3.B) Identification of binding partners. We have performed several state-of-the-art techniques to identify binding partners for the candidate lncRNAs. 1. Interacting proteins have been identified by RNA immunoprecipitation (RIP) and in vitro binding assays. 2. DNA and RNA interacting molecules have been identified using the novel technique Chromatin Isolation by RNA purification (ChIRP) using biotinylated antisense DNA oligos. For the first part of the project ChIRP was followed by qPCR to detect interactions with the DNA from a specific locus. For the second part of the project, in order to identify genome wide RNA:RNA interactions, the samples have been sent for sequencing to BGI, China. 3. Genomic interactions between different chromatin locations have been identified by Circularize Chromosome Conformation Capture followed by sequencing (4C-sequencing). I performed the experiment in triplicates and the samples were sequencing at BGI in China. Dr Harmen J.G. van de Werken at the Erasmus Medical Center in Rotterdam carried out the bioinformatic analysis. 3. A description of the main results achieved so far: Using the data from the RNA-sequencing and the microarray platform we have identified the lncRNA MIR31HG to be important for oncogenic-induced senescence. MIR31HG is upregulated during senescence and it is located approximately at 400 Kilobases from the INK4A locus, key regulator of cellular senescence. In human primary fibroblasts, MIR31HG exists in a unique isoform of approximately 2.1 Kilobases long. In proliferating fibroblasts, MIR31HG is expressed at basal levels (approximately 40 copies per cell) and located both in the nucleus and the cytoplasm, whereas during OIS its expression is highly increased reaching 250 copies, which are mainly cytoplasmic. - In the first period of the project we focused in the nuclear function of MIR31HG in proliferating cells. We observed that depletion by siRNAs of this lncRNA turned the cells into senescence. Following siRNA depletion we observed that the cells reduce the cell growth rate and acquire other senescence features such as increased cell cycle inhibitor p16INK4A protein and mRNA, activation of Senescence Associated-β-galactosidase activity enzyme and the presence of senescence-associated heterochromatin foci. The phenotype observed was fully rescued by inhibiting p16INK4A protein using siRNAs meaning that the effect of MIR31HG knockdown in senescence was due to the increase in p16INK4A protein levels. We have elucidated the molecular mechanism by which MIR31HG is regulating senescence. We have showed that MIR31HG interacts with INK4A promoter and with Polycomb group (PcG) proteins, being required for PcG repression of the INK4A locus. We have also found interactions between MIR31HG promoter and an enhancer that is highly active upon BRAF expression, suggesting that MIR31HG expression might be driven by this enhancer during OIS. -In the second period of the project, we focused on the cytoplasmic function during OIS. During OIS the cells secrete a number of cytokines and interleukins to the media that could impact the neighboring cells in different manner. These secreted factors are the so-called senescence-associated secretory phenotype (SASP). Although some factors can have a good effect reinforcing senescence and regenerating damaged tissues, other pro-inflammatory factors can impact on a negative manner the surrounding tissues by creating an inflammatory environment that could benefit tumor progression and invasion. We have observed that depletion of MIR31HG in cells undergoing OIS reduces the expression of the SASP components, both at the RNA level in the cell but also as secreted factors to the media. We have assessed the impact of this SASP by taking the conditioned media (CM) from control-senescent cells and MIR31HG knockdown-senescent cells and adding it to primary fibroblasts or cancer cells. Fibroblast receiving the CM from control and MIR31HG knockdown-senescent cells are able to undergo senescence suggesting that the SASP components that the reinforce senescence are present on both SASP. Breast cancer cells (MDA-MB-231) that are in contact with the CM from MIR31HG knockdown-senescent cells are less prone to invade a matrigel basement membrane compare to cells exposed to CM from control-senescent cells suggesting that factors promoting invasion are reduced in the CM from MIR31HG knockdown-senescent cells. We are currently working on elucidating the molecular mechanism underlying the SASP regulation. 4. The expected final results and their potential impact and use: - For the first part of the project we have identified a new regulatory molecule of important the cell cycle inhibitor and tumor suppressor p16INK4A. This molecule, the lncRNA MIR31HG is upregulated upon oncogenic BRAF expression. Overexpression of the BRAF oncogene due to constitutively active mutations is frequently observed in melanoma cancer. We have observed a negatively correlation between MIR31HG and p16INK4A mRNA expressions in melanoma patients bearing this type of mutations indicating a physiological relevance for this MIR31HG in melanoma. The results of this work have been published in the scientific journal Nature Communications. - For the second part of the project, our results suggest that the SASP generated from MIR31HG down-senescent cells is different in composition to control-senescent cells. The positive side of the SASP seems to remain in the CM from MIR31HG knockdown- senescent cells whereas the bad effect, due to pro-inflammatory factors is reduced. These preliminary results might help to elucidate the classification of the “good and the bad” components of the SASP and make MIR31HG very attractive as a therapeutic use of these senescent cells. Some information about the project can be found at the Lund group website http://www.bric.ku.dk/research/lund_group/.

Data: CORDIS, © European Union