High-through-put Screening And Functional Study Of Non-coding RNAs And Their Regulatory Roles In The Self-renewal Of Human Embryonic Stem Cells

Background and objectiveEmbryonic stem cells (ESC) derived from the inner cell mass, which possess thepluripotency for unlimited proliferation and differentiation. Pluripotency is defined, in abroad sense, as the ability of a cell to give rise to derivatives of the three germ layers. Inthe spectrum of cell potency, totipotency represents the cell with the greatestdifferentiation potential, represents the ability of a single cell to divide and produce all ofthe differentiated cells or into an individual body. Pluripotency, on the other hand, refersto stem cells that have the potential to differentiate into any of the three germ layers.However, cell pluripotency is a continuum, there are cells exhibit complete pluripotentthat can form every cell of the embryo (e.g. embyronic stem cells), also there are cellsthat exhibit incomplete or partially pluripotent that can form cells of all three germ layersbut that may not exhibit all the characteristics of completely pluripotent cells. Andmultipotency cells are progenitor cells which have the gene activation potential todifferentiate into multiple, but limited cell types. The Embryonic stem cells we arestudying harbors the strongest ‘stemness’ of all pluripotency cells, and are an ideal toolfor clinical use such as tissue engineering or regenerative medicine. However, To fullyexplore their potential requires a deeper understanding of the molecular basis ofstemness.ESC pluripotency is tightly controlled. Amongst several early attempts in revealingsignaling pathways that control ESC pluripotency, the Nodal/Activin branch and the thebone morphogenetic protein (BMP) branch are most studied. Downstream of thosesignaling pathways, the maintenance of ESCs pluripotency is ensured by a regulatorycircuitry including three main core transcription factors (TFs), Oct4, Sox2and Nanog.The seminal work done by Shinya Yamanaka showed that pluripotency could be inducedin mouse somatic cells by ectopic expression of those core transcription factors, whicharoused the interest of many scientists to thoroughly study the regulation network of thesecore transcription factors. Recently, new players like non-coding RNAs (ncRNAs) joinedthe web of pluripotency network.Non-coding RNA molecules, transcribed by RNA polymerase II, are previouslyregarded to exert only passive roles in the cell. But till now, many are defined conversely as primary players in determining cell development, metabolism, differentiation andhomeostasis. Among which, small non-coding RNAs (<200nucleotides), especiallymicroRNAs (miRNAs) are now considered major regulators of in all multicellularorganisms. However long non-coding RNAs (>200nucleotides, lncRNAs) are seldomstudied in hESC, especially in the regulation of pluripotency. The recent work done byour group showed that lincRNA-ROR functions as a sponge to alleviate the expression ofcore transcription factors from miRNAs binding. Beside this many study showed thatlncRNA can activate transcription, cause epigenetic modification, and have a role inposttranscriptional regulation. Taken all, the emerging roles of non-coding RNAs in theregulation of pluripotency network is a field that worth studying.Previous studies mainly focused on a single ncRNA regulating a set of genes. As ourknowledge of the transcriptome space has expanded, it has become increasingly clear thatnumerous miRNA-binding sites exist on a wide variety of RNA transcripts, leading to thehypothesis that all RNA transcripts that contain miRNA-binding sites can communicatewith and regulate each other by competing specifically for shared miRNAs, thus acting ascompeting endogenous RNAs (ceRNAs). Any element that consists this miRNA-RNAnetwork are essential in keeping the system in harmony. With perturbation of a singlenode, the whole network rearranged, causing a series of change in gene expression thatmay even change the phenotype of a cell.However, in ESC the same network exists. Our previous work served as a goodexample by demonstrating lincROR sponged mir-145to increase pluripotency. Our studyin this paper describes another competing endogenous network controlled by ncRNAs.We first studied miRNAs that promote differentiation of hESC by targeting coretranscription factors Oct4，Nanog and Sox2. After that, we identified lncRNA-GAS5, agrowth arrest associated lncRNA, strongly correlates with the pluripotency. Functionalstudies showed that GAS5attenuates miRNA that target pluripotency related TGFβreceptor ligand NODAL, thus promoting hESC self-renewal and pluripotency. Altogether,we demonstrated a new pluripotency regulatory network through competing endogenousmechanism.Part ⅠThe effect of pluripotency targeting miRNAs in hESCMethods:(1) Searching NCBI GEOdatabase for a miRNA dataset for furtheranalysis.(2) Using a targeting site prediction software to find potential differentiation promoting miRNAs by predict possible bindings to the mRNA of Oct4, Nanog and Sox2.(3) Dual-Luciferase reporter assay and overexpression of these candidates wereperformed; mRNA and protein levels of Oct4, Nanog and Sox2were analyzed.(4)Evaluate the expression of mir-149and mir-320in an induced differentiation model. APstaining and immunocytochemistry were performed to assess their effect on self-renewaland pluripotency.(5) Overexpress or inhibit mir-149or mir-320and study thedifferentiation pattern by real-time PCR.Results:(1) We searched Gene Expression Omnibus (GEO) data base, and used amiRNA dataset (GSE14473) containing8different human ESC (hESC) lines for furtheranalysis. we clustered the log transferred data of differentiated or undifferentiated miRNAexpression in all10cell lines using unsupervised, hierarchical clustering, and picked66up-regulated miRNAs that clustered at a relatively high distance (D=0.668) as putativemiRNAs.(2) We further narrowed down the candidates using a miRNA target predictionalgorithm miRanda by predicting possible bindings to the mRNA of Oct4, Nanog andSox2. Six candidates were selected for further analysis.(3) The results showed thatmir-149and mir-320strongly attenuates the expression of Oct4and causing hESC todifferentiate.(4) We found that mir-149and mir-320were strongly up-regulated uponhESC differentiation, which exerts a negative relation with the expression of pluripotencyfactors.(5) All ectodermal lineage markers up-regulated during mir-149and mir-320overexpression, and attenuated during inhibition. When compared with that of Oc4knockdown, this expression pattern seems intriguing. By functional analysis, weidentified another important gene, β-catenin, is targeted by mir-149and mir-320, and isthe cause for such lineage shift.Conclusions: We employed bioinformatics and computational attempts to identifypotential differentiation promoting miRNA in the chip data. By a series of functionalanalysis, we identified mir-149and mir-320can target Oct4and interrupt the self-renewalof hESC. In studying the differentiation lineage pattern of mir-149and mir-320, weidentified another hESC regulation key pathway gene, β-catenin, which is responsible forthe ectodermal promotion phenotype produced by overexpression of mir-149andmir-320. Part ⅡLncRNA-GAS5in maintains pluripotency of hESCMethods:(1) To investigate LncRNA that controls hESC pluripotency, weperformed RNA-sequence to identify LncRNAs that has difference expression betweendifferentiation and ground state.(2) We constructed several overexpression plasmids ofcandidate lncRNAs and transfected them into hESC to seek phenotype changes.(3) Wegenerated stable-expressed ES cell line by lentiviral infection. After that we assessed thephenotypic change.(4) We performed cell counting experiment and assessed the cellcycle change using either GAS5overexpressed or knockdown cell line. Also we assessedthe pluripotency of these cell lines using AP staining and Realtime PCR.(5) In situhybridization and nuclear-cytoplasm separation were performed to assess the localizationof GAS5.Results:(1) We performed RNA-sequence of two ES cell lines (X-01and H1) intwo states (undifferentiated and differentiated), and picked lncRNAs that expressedmostly in the undifferentiated state for further analysis.(2) We constructedover-expression plasmid of DANCR、LOC100506647、LINC00458、GAS5, and test theirfunction by transfection into hESC. We found that GAS5strongly elevated the expressionof Oct4, Nanog and Sox2.(3) Cell lines that stably express GAS5or GAS5-siRNA weregenerated using lentiviral infection and puromycin selection.(4) Experiments showedthat overexpression of GAS5significantly improved the proliferation of hESC, increasedS phase and reduced G1phase cells were observed. AP staining showed that hESCcolonies were increased in number when overexpressing GAS5and number dropped inknockdown group.(4) In situ hybridization and nuclear-cytoplasm separation experimentshowed that GAS5mainly expressed in cytoplasm, and decreased significantly uponhESC differentiation. Indicating GAS5expression is strongly related to the mRNA levelsof Oct4and Sox2.Conclusions: Using high-through-put technology, we identified a set of lncRNA thatcorrelates with hESC self-renewal. By functional analysis, we showed that GAS5canelevate the expression of Oct4, Nanog and Sox2, and increase the cell cycle of hESC.Localization experiments showed that GAS5mainly stayed in cytoplasm, giving cues thatit may regulates the pluripotency in a posttranscriptional fashion.Part Ⅲ GAS5increases pluripotency by competing endogenous miRNAsMethods:(1) Identify genes that correlate with GAS5expression by usinghigh-through RNA-seq.(2) RNA-IP of Ago2was performed to see if GAS5perturbation affected the miRNA binding on the mRNA of NODAL. ShRNA that targets Dicerwith/without GAS5overexpression was performed using lentiviral approach. Theexpression change was assessed by real-time PCR.(3) Using miRNA-seq technology toidentify miRNAs that correlate negatively with GAS5, NODAL (4) To find whetherpromoter of GAS5is bound by OCT4or SOX2, we performed chIP assay, and simply byoverexpressing the protein of NODAL.Results:(1) By using high-through RNA-seq, we identified a set of genes thatcorrelates with GAS5expression, including pluripotency related pathway regulatorNODAL and TGFB3, also the G1phase modulator CDK4. After analysis of severalsignaling pathways, we indentified NODAL-TGFβ-SMAD2/3to be the most functionalpathway that relates to GAS5overexpression.(2)RNA-IP of Ago2gave strong evidencethat overexpression of GAS5attenuated the miRNA binding on the mRNA of Oct4andSox2. Overexpression of GAS5failed to increase OCT4and SOX2expression in Dicerknockdown hESC, indicating this effect is partially mediated through miRNA.(3) Bydata mining, we identified56putative miRNA candidates that target GAS5and NODAL.We selected the top10miRNA candidates, and expressed them using miRNA mimics.5of them showed strongly down-regulated the expression of GAS5and NODAL.(4) ChIPassay showed that the core promoter region is bound by both OCT4and SOX2, andoverexpressing OCT4and SOX2cause strong up-regulation of GAS5.Conclusions: Here we showed that GAS5increases the pluripotency of hESC in amiRNA dependent fashion. Using miRNA-seq technology we identified miRNAs thatparticipate in the GAS5-NODAL-pluripotency regulation network.ConclusionsIn summary, the work presented here elucidates a new layer of pluripotencyregulation network in hESC. we identify that GAS5(growth arrest-specific transcript5),a tumor suppressor and growth arrest gene which is abundantly expressed in thecytoplasm of hESCs, accelerates the G1/S phase transition and promotes stemness ofhESCs by acting as a competing endogenous RNA (ceRNA). By functional analysis, weshow that GAS5regulates the expression of core signaling transduction factor NODALthrough a miRNA-dependent fashion. We conclude that GAS5is of great importance tothe self-renewal of hESCs by acting as a ceRNA. This serves as a good example of how alncRNA’s function may vary in different biological context, and gives us the idea of how miRNome can change a gene’s function. Our findings thereby offer a unique perspectivefor understanding the mechanisms of pluripotency maintaining of ESCs in the human.