Clarifying the regulatory mechanisms of embryonic stem cell (ESC) neural differentiation is helpful not only for understanding neural development but also for obtaining high‐quality neural progenitor cells required by stem cell therapy of neurodegenerative diseases. In this article, Kang and coworkers show that lncRNA‐1604 is required for the neural differentiation process of mouse ESCs, and functions as a ceRNA of miR‐200c and regulates ZEB1/2 to orchestrate neural differentiation. These findings not only identify an unrevealed role of lncRNA‐1604 and ZEB1/2 but also elucidate a new regulatory pathway of lncRNA‐1604/miR‐200c/ZEB axis in neural differentiation.
Neural differentiation is a complex and precisely regulated process during early development of the central nervous system. Accumulating studies have indicated that abnormal neural differentiation is associated with several neurological and psychiatric disorders, and seriously compromises human health and quality of life 1. Establishment of embryonic stem cell (ESC) neural differentiation systems in vitro has provided insight into the mechanisms underlying the nervous system development and associated diseases, and also provided a rich source of neural progenitor cells for the study of stem cell therapy for neurodegenerative diseases 2-5. It has been reported that the neural differentiation potential of mouse ESCs can be affected by the intrinsic multilayer regulatory network, including core transcription factors, epigenetic modifications, microRNAs (miRNAs), and signaling pathways 6. However, the function and mechanism of long noncoding RNAs (lncRNAs) in neural differentiation of mouse ESCs remains to be elucidated.
LncRNAs as a group of RNAs (>200 nucleotides) with little or no protein coding potential 7 can participate in various biological processes by regulating gene expression through epigenetic modification 8, and controlling transcription or translation 9, 10. Recently, systematic loss‐of‐function studies of most lncRNAs known to be expressed in mouse ESCs indicated that knockdown of dozens of lncRNAs caused either exit from the pluripotent state or upregulation of lineage commitment programs 11. It has been reported that lncRNAs may not only contribute to pluripotency acquisition and maintenance (such as lncRNA‐ROR) 12 but also contribute to differentiation of multiple lineages including muscle 13, 14 and neuron 15. The present study showed that lncRNA‐1604 (also termed as GM11627, chr11:102576443–102579275) was highly expressed in cytoplasm during neural differentiation, but no literature is available on the function of lncRNA‐1604 in neural differentiation of mouse ESCs.
Generally, lncRNAs act as decoys, guides, and scaffolds that interfere with or enhance the function of other regulators including protein and miRNAs by directly interacting with DNA, RNA, or proteins 16. In ESCs, lncRNAs can control pluripotency and multiple lineage differentiation by interacting with epigenetic modifiers, RNA‐binding proteins, and miRNAs 12. Accumulating evidence has revealed that lncRNAs can regulate gene expression either in cis or in trans in the nucleus 16, and act as competitors or reservoirs of miRNAs to indirectly regulate gene expression in the cytoplasm 17, 18. Several cytosolic lncRNAs have been shown to predominantly function as competing endogenous RNAs (ceRNAs) to modulate the activities and biological functions of miRNAs. The lncRNA‐MD1 sponges miR‐133 and miR‐135 to protect the mRNA of the muscle‐specific genes MAML1 and MEF2C 13, and lncRNA‐RoR functions as a ceRNA and sequesters miR‐145 to regulate the core transcription factors Oct4, Sox2, and Nanog in human ESCs 19. Through bioinformatics prediction analysis of the binding energy between lncRNA‐1604 and miRNAs, we found that lncRNA‐1604 may be a significant ceRNA of miR‐200c. MiR‐200c plays a critical role in modulating the mesenchymal‐to‐epithelial transition in somatic cell reprogramming 20-22 and cell fate decision of human ESC differentiation 23. However, whether lncRNA‐1604 functions as a ceRNA and sequesters miR‐200c to regulate neural differentiation of mouse ESCs remains unclear.
In this study, we found that lncRNA‐1604 orchestrated neural differentiation through the miR‐200c/ZEB axis, providing new evidence for the epigenetic regulatory mechanisms of neural differentiation from mouse ESCs.
Mouse Sox1‐green fluorescent protein (GFP) ESCs (46C) 3, as a gift from A. Smith, and D3 cells were maintained on feeders in mESC culture conditions. For neural differentiation, ESCs were suspension cultured in differentiation medium [Glasgow's Minimum Essential Medium supplemented with 8% Knockout Serum Replacement (KOSR), 2 mM glutamine, 1 mM sodium pyruvate, 0.1 mM non‐essential amino acids, and 0.1 mM 2‐mercaptoethanol with or without 5% fetal bovine serum (FBS)]. The fluorescence‐activated cell sorter (FACS) analysis was immediately performed to detect the percentage of Sox1‐GFP positive (Sox1‐GFP ) cells after dissociated to single cells.
For generating vectors of sh1604‐1/2, shZEB1‐1/2, and shZEB2‐1/2, the specific 21 nucleotides targeting the transcripts of lncRNA‐1604, ZEB1, or ZEB2 were designed (sh1604‐1, 5′‐TACCCAAATAGAACGTCATTT‐3′; sh1604‐2, 5′‐GCTGTACCCAA ATAGAACGTC‐3′; shZEB1‐1, 5′‐ATAGAGGCTACAAGCGCTTTA‐3′; shZEB1–2, 5′‐CCGCCAACAAGCAGACTATTC‐3′; shZEB2‐1, 5′‐CCACTAGACTTCAATGACTAT‐3′; and shZEB2‐2, 5′‐CCGAATGAGAAACAATATCAA‐3′) and cloned into the pLKO.1‐TRC vector for RNA interference. The full‐length transcripts of lncRNA‐1604, ZEB1, and ZEB2 were respectively cloned into the FUW vector with the corresponding primers (Supporting Information Table S1). DNA fragment corresponding to the pre‐miR‐200c was amplified by polymerase chain reaction (PCR) from mouse genomic DNA and cloned into the FUW vector.
Total RNA was extracted using RNAiso Plus (TaKaRa, Otsu, Japan), and the first strand of cDNA synthesis was performed using the PrimeScript RT reagent kit (TaKaRa, Otsu, Japan), according to the manufacturer's instructions. Quantitative reverse transcriptase PCR (RT‐PCR) analysis was carried out using the SYBR Green quantitative PCR (qPCR) Master Mix (Bio‐Rad, CA, USA). MiRNA levels were measured using the Bulge‐Loop miRNA qPCR Primer Set (RiboBio, Guangzhou, China) according to the manufacturer's instructions. The relative expression level of each candidate gene was calculated by the 2–ΔΔCt method 24, using GAPDH and U6 as the internal normalized control. Each experiment was performed independently and in triplicate. The corresponding quantitative RT‐PCR primers were shown in Supporting Information Table S2.
Nuclear and cytoplasmic RNA was isolated as described previously 25. Mouse ESCs were collected and washed twice with ice‐cold phosphate‐buffered saline (PBS) and centrifuged at 1,000 rpm for 10 minutes. Cell pellets were resuspended gently and incubated on ice for 5 minutes in 200 µl lysis buffer A [Tris (10 mM, pH 8.0), NaCl (140 mM), MgCl2 (1.5 mM) 0.5% Nonidet P‐40 (NP‐40)] and then centrifuged at 1,000 g for 3 minutes at 4°C. The supernatant with the cytoplasmic fraction was added to 1 ml RNAiso Plus. Nuclear pellets were washed and then resuspended in 1 ml RNAiso Plus. Then, purification and analysis of cytoplasmic and nuclear RNA were performed by quantitative RT‐PCR.
RNA sequencing library generation, workflow, data analysis, and enrichment analysis were performed as reported previously 26. In brief, the RNA sequencing library was generated according to the methods described previously 27 and sequencing was performed by illumina Hiseq2500/Hiseq3000 platform. After trimmed using sickle.pe (pair‐end) (v1.29, https://github.com/najoshi/sickle) with the parameters (–q 20, –l 30), the RNA sequencing reads were mapped to the mouse genome (mm10) using Tophat (2.0.7) with the default parameters and Ensemble genome annotation (Mus_musculus.GRCm38.73.gtf) 28. The expression level (fragments per kilobase of exon per million fragments mapped) of each gene was estimated using Cufflinks (v2.0.2) software 29, while the differentially expressed genes (DEGs) were detected using Cuffdiff 30. Multiple tests were adjusted using false discovery rate (FDR) methods, and a cutoff FDR < 0.05 was chosen for statistical significance. The gene ontology (GO) enrichment analyses were completed using DAVID web servers 31 and results were presented as rich factor plots.
The MS2bp‐MS2bs‐based RNA immunoprecipitation (RIP) assay was performed as described previously 19, 32. We constructed vectors expressing lncRNA‐1604 combined with MS2bs elements and cotransfected miR‐200c mimics and MS2bp‐YFP vector into HEK293 cells with FuGENE HD reagent. After 48 hours, cells were collected and subsequently lysed with lysis buffer [KCl (100 mM), MgCl2 (5 mM), HEPES (10 mM, pH 7.0), NP‐40 (0.5% v/v), DL‐Dithiothreitol (1 mM), propidium iodide (100×), RNaseOut (0.1 U/µl), phenylmethanesulfonyl fluoride (1 mM)] for 30 minutes on ice. The transcript‐specific binding RNA‐protein complexes were then immunoprecipitated with YFP antibody (Abcam, Cambridge, MA, USA) or control rabbit IgG (Cell Signaling Technology, Danvers, MA, USA). The RNA complexes were isolated using phenol–chloroform extraction and analyzed by quantitative RT‐PCR.
Luciferase reporter vectors were generated by cloning lncRNA‐1604 or mutant lncRNA‐1604 for miR‐200c targeting sites into pGL3‐based vectors. Mutant vectors were obtained using the QuikChange Lightning Multi Site‐Directed Mutagenesis Kit (Agilent Technologies, Santa Clara, CA, USA). The luciferase reporter vectors of pGL3‐ZEB1 and pGL3‐ZEB2 have been reported previously 21. NIH3T3 cells (3 × 104) were transfected with 100 ng of luciferase reporter vector, 10 ng of Renilla luciferase vector (pRL‐TK), 50 nM micrON miRNA mimics, or 150 nM micrOFF miRNA inhibitors synthesized by Ribobio (Guangzhou, China) with 1 µl of FuGENE HD reagent. Firefly and Renilla luciferase activities in cell lysates were assayed with a Dual‐Luciferase Reporter Assay System (Promega, Madison, WI, USA) at 48 hours post‐transfection. Firefly luciferase activity was normalized by Renilla luciferase.
Cells were collected and lysed on ice for 30 minutes with Radio Immunoprecipitation Assay lysis buffer (Beyotime, Shanghai, China). Protein concentrations were measured using the Pierce BCA Protein Assay Kit (Thermo, Waltham, MA, USA). Cell lysate was separated using SDS‐polyacrylamide gel electrophoresis, blotted on poly(vinylidene fluoride) membranes, and probed with the following primary antibodies: ZEB1 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), ZEB2/SIP1 (Novus, Colorado, USA), E‐cad (Cell Signaling Technology, Danvers, MA, USA), and GAPDH (Biogot Technology, Nanjing, China). GAPDH was used as the loading control. After incubation with secondary antibodies, signals were visualized by enhanced chemiluminescence.
Cells were washed with PBS, fixed with 4% paraformaldehyde (PFA) for 20 minutes, and then permeabilized with 0.2% Triton X‐100 for 8 minutes. Cells were then washed twice with PBS, blocked in 10% FBS for 45 minutes and incubated overnight with primary antibodies at 4°C. Then, the cells were stained with fluorescent secondary antibodies for 2 hours and Hoechst 33342 (Ho.33342) for 20 minutes. Images were acquired using a fluorescence microscope. Primary antibodies included Sox1 (Abcam, Cambridge, MA, USA), N‐cad (Cell Signaling Technology, Danvers, MA, USA), ZEB1 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), and ZEB2/SIP1 (Novus, Colorado, USA).
To generate teratomas, ESCs were trypsinized, resuspended at a concentration of 1.5 × 106 cells per 150 μl, and then injected into athymus nude mice. Tumors were assessed every week for 4 weeks, harvested, and fixed in 4% PFA for 24 hours at room temperature before paraffin embedding, and then stained using H&E. All experiments were approved by the Institutional Animal Care and Use Committee of Tongji University.
The statistical significance was analyzed by one‐way analysis of variance and Student t test (two‐tailed). All error bars represent the SEM from three independent experiments. *, #, p < .05; **, ##, p < .01; ***, ###, p < .001.
Through quantitative RT‐PCR analyses, we found that the expression level of lncRNA‐1604 was significantly higher in mouse ESCs than mouse embryonic fibroblasts (Fig. 1A). For mouse tissues, lncRNA‐1604 was predominantly detected in the brain and striatum, but was virtually absent in the other tissues (Fig. 1B). These findings indicated that lncRNA‐1604 might be a specific and critical regulator of neural differentiation. Furthermore, gene expression analysis of the nuclear and cytoplasm fractions showed that lncRNA‐1604 was primarily expressed in the cytoplasm (Fig. 1C). To investigate the specific function of lncRNA‐1604 in pluripotency and differentiation potential of mouse ESCs, we constructed lentiviral vectors (sh1604‐1 and sh1604‐2) including the short hairpin RNA (shRNA) targeting lncRNA‐1604 (Gm11627‐004). The knockdown efficiency of sh1604‐1 and sh1604‐2 on lncRNA‐1604 was confirmed by quantitative RT‐PCR analysis (Fig. 1D). Then, we found that knockdown of lncRNA‐1604 showed no obvious effect on cell morphology and pluripotency‐associated gene expression of ESCs (Supporting Information Fig. S1).
Knockdown of lncRNA‐1604 inhibits ectoderm differentiation potentials. (A): The expression level of lncRNA‐1604 in mouse embryonic fibroblast and mouse embryonic stem cells (ESCs). (B): The expression level of lncRNA‐1604 in mouse tissues including gut, liver, muscle, ovary, kidney, brain, striatum, and cerebellum. (C): The distribution of lncRNA‐1604 in the cytoplasm and nucleus of mouse ESCs. (D): Quantitative reverse transcriptase polymerase chain reaction (qRT‐PCR) analysis of the knockdown efficiency of shRNA targeting lncRNA‐1604 (sh1604‐1 and sh1604‐2) compared with shCtrl. (E): Effects of lncRNA‐1604 knockdown on the morphology and weight of teratoma. (F): qRT‐PCR analysis of the expression level of ectoderm and mesendoderm genes in teratomas derived from lncRNA‐1604 knockdown (sh1604‐1 and sh1604‐2) compared with shCtrl. (G): The H&E staining assay for the three types of lineage of teratomas derived from lncRNA‐1604 knockdown (sh1604‐1 and sh1604‐2) and control (shCtrl) mouse ESCs. (H): Immunostaining assay for N‐cad in teratomas derived from lncRNA‐1604 knockdown (sh1604‐1 and sh1604‐2) compared with shCtrl. Ho.33342 was used for nucleus staining (blue). GAPDH was used as the internal control. Scale bar = 100 μm. Error bars showed SEM (n = 3). *, p < .05; **, p < .01; ***, p < .001. Abbreviations: ESC, embryonic stem cell; MEF, mouse embryonic fibroblast; shCtrl, control.
To clarify the effects of lncRNA‐1604 on the differentiation potentials of mouse ESCs, we carried out teratoma formation experiments in vivo with athymic nude mice. Results showed that there was no significant difference in morphology or weight of teratomas derived from lncRNA‐1604 knockdown (sh1604‐1 and sh1604‐2) ESCs compared with shCtrl (Fig. 1E). However, we found that the teratomas generated from lncRNA‐1604 knockdown (sh1604‐1 and sh1604‐2) ESCs presented significantly lower mRNA levels of neural‐lineage genes (Sox1, Nestin, Zfp521, Pax6, N‐cad, and GFAP), while the hallmark mesendoderm lineage genes were partially upregulated compared with shCtrl (Fig. 1F). These findings were confirmed by histological analysis of the three germ layers, such as minimal neural tube‐like structures (Fig. 1G), and immunostaining with about 76% decrease for the median intensity of N‐cad (Fig. 1H), which demonstrated that knockdown of lncRNA‐1604 inhibits ectoderm differentiation of mouse ESCs.
To explore the biological function of lncRNA‐1604, we performed RNA sequencing assay to obtain the DEGs after knockdown or overexpression of lncRNA‐1604 in mouse ESCs. We found that there were 139 genes upregulated and 97 genes downregulated after lncRNA‐1604 knockdown (sh1604‐1) compared with shCtrl (Fig. 2A; Supporting Information Table S3), while there were 156 genes upregulated and 241 genes downregulated after lncRNA‐1604 overexpression (FUW‐1604) compared with FUW‐Luc (Fig. 2B; Supporting Information Table S3). For the 34 DEGs consistently regulated by lncRNA‐1604 knockdown and overexpression (Supporting Information Table S3), we carried out GO term enrichment analysis. Results showed that these DEGs mainly focused on neural development (Fig. 2C), indicating that lncRNA‐1604 may play critical function in the neural differentiation of mouse ESCs.
LncRNA‐1604 is required for neural differentiation process of mouse embryonic stem cells (ESCs). (A): A volcano plot of the differentially expressed genes (DEGs) between lncRNA‐1604 knockdown (sh1604‐1) and control (shCtrl) mouse ESCs. The x‐axes index the log2FC (sh1604‐1 vs. shCtrl), and the y‐axes index the –log10(p value). (B): A volcano plot of the DEGs between lncRNA‐1604 overexpression (FUW‐1604) and control (FUW‐Luc) mouse ESCs. The x‐axes index the log2FC (FUW‐1604 vs. FUW‐Luc), and the y‐axes index the –log10 (p value). (C): GO analysis for the biological function of these 34 consistent DEGs regulated by lncRNA‐1604. The x‐axes index the rich factor, and the y‐axes index GO terms. (D): The representative Sox1‐green fluorescent protein (GFP) positive (Sox1‐GFP ) images for the effects of lncRNA‐1604 knockdown on the neural differentiation process of mouse ESCs. (E): The fluorescence‐activated cell sorter assay of Sox1‐GFP cells for the effects of lncRNA‐1604 knockdown on the neural differentiation process of mouse ESCs compared with shCtrl. (F): Immunostaining assay for the representative genes Sox1 and N‐cad in neural progenitor cells derived from lncRNA‐1604 knockdown (sh1604‐1 and sh1604‐2) and control (shCtrl) mouse ESCs. Ho.33342 was used for nucleus staining (blue). (G): Quantitative reverse transcriptase polymerase chain reaction analysis for the expression level of neural progenitor cell markers (Sox1, Nestin, Zfp521, Pax6, and N‐cad) during the neural differentiation process of mouse ESCs after lncRNA‐1604 knockdown compared with shCtrl. GAPDH was used as the internal control. Scale bar = 100 μm. Error bars showed SEM (n = 3). *, p < .05; **, p < .01; ***, p < .001. Abbreviations: FC, fold change; GO, gene ontology; GFP, green fluorescent protein.
To observe the effects of lncRNA‐1604 on the neural differentiation potentials of mouse ESCs, we performed neural progenitor differentiation with 8% KOSR medium as described previously 33. The results showed that lncRNA‐1604 was continuously and highly expressed in the neural differentiation process of mouse ESCs (Supporting Information Fig. S2A). After knockdown of lncRNA‐1604 (sh1604‐1 and sh1604‐2), we found that the percentage of Sox1‐GFP positive (Sox1‐GFP ) cells was significantly decreased at days 3, 5, and 7 compared with that of shCtrl (Fig. 2D, 2E). Consistently, the immunostaining assays also confirmed the expression of Sox1 and N‐cad in neural progenitor cells generated from ESCs following knockdown of lncRNA‐1604 (sh1604‐1 and sh1604‐2) (Fig. 2F). The mRNA levels of neural progenitor cell markers (Sox1, Nestin, Zfp521, Pax6, and N‐cad) were significantly decreased upon knockdown of lncRNA‐1604 during the neural differentiation process (Fig. 2G). These findings were confirmed in another mouse ESC line (D3) (Supporting Information Fig. S2B, S2C) and lncRNA‐1604 knockout (Supporting Information Fig. S2D, S2E), further demonstrated that lncRNA‐1604 is required for the neural differentiation of mouse ESCs.
To explore the functional mechanism of lncRNA‐1604, we performed the bioinformatics prediction analysis to assess whether lncRNA‐1604 could act as ceRNA. Results showed that lncRNA‐1604 harbors many miRNAs targeting sites, including miR‐200c and miR‐429, two pivotal members of miR‐200 family. Further analysis for the binding energy between lncRNA‐1604 and miRNAs indicated that miR‐200c‐5p and miR‐200c‐3p were the strongest candidates (Fig. 3A).
LncRNA‐1604 serves as a competitor of miR‐200c to regulate ZEB1/2. (A): The binding energy between lncRNA‐1604 and miRNAs (miR‐200c‐5p, miR‐200c‐3p, miR‐429‐5p, and miR‐429‐3p). (B): Dual‐luciferase reporter gene assays for the interaction between lncRNA‐1604 and miRNAs (miR‐200a, miR‐200b, miR‐200c, miR‐141, and miR‐429) with co‐transfection of pGL3‐lncRNA‐1604, pGL3‐Renilla, and the corresponding miRNA mimics. (C): Dual‐luciferase reporter gene assays for the effects of the miR‐200c inhibitor on that of the miR‐200c mimics on the luciferase activity of pGL3‐lncRNA‐1604. (D): Dual‐luciferase reporter gene assays for the effects of miR‐200c mimics on the luciferase activity of pGL3‐lncRNA‐1604 vector with miR‐200c binding sites mutant (pGL3‐lncRNA‐1604‐mutant). (E): Quantitative reverse transcriptase polymerase chain reaction (qRT‐PCR) analysis for the RIP RNAs to detect the enrichment of miR‐200c on lncRNA‐1604 after co‐transfection of MS2bs‐1604, MS2bp‐YFP, and miR‐200c mimics, compared with MS2bs‐1604. (F): Dual‐luciferase reporter gene assays for the effects of lncRNA‐1604 overexpression (FUW‐1604) on that of miR‐200c mimics on the luciferase activity of pGL3‐ZEB1 and pGL3‐ZEB2. (G): Western blotting analyses for the effects of lncRNA‐1604 knockdown (sh1604‐1) on the expression level of ZEB1 and ZEB2. (H): Immunostaining assays for the expression level of ZEB1 and ZEB2 in teratomas derived from lncRNA‐1604 knockdown (sh1604‐1 and sh1604‐2) compared with shCtrl. Ho.33342 was used for nucleus staining (blue). (I): qRT‐PCR and Western blotting analysis for the effects of lncRNA‐1604 overexpression (FUW‐1604) on the repressive effects of miR‐200c overexpression (FUW‐miR‐200c) on the expression level of ZEB1 and ZEB2. (J): Western blotting analysis for the effects of miR‐200c sponge on lncRNA‐1604 knockdown (sh1604‐1) induced ZEB1 and ZEB2 downregulation. Scale bar = 100 μm. GAPDH was used as the loading control. GAPDH and U6 were used as the internal control for mRNA and miRNA, respectively. Error bars showed SEM (n = 3). The asterisk (*) denotes a significance versus “mc,” “MS2bs‐RL,” “FUW‐Luc,” or “shCtrl”. The hash mark (#) denotes a significance versus “miR‐200c FUW,” “FUW‐miR‐200c,” or “sh1604‐1 FUW‐Luc”. *, #, p < .05; **, p < .01; ***, p < .001. Abbreviation: mfe, minor free energy; mc, mimics control; miR‐200c i, miR‐200c mimics and inhibitor.
To confirm that lncRNA‐1604 was a functional ceRNA of miR‐200c, we performed dual‐luciferase reporter assays to confirm their direct interaction. Results showed that introduction of miR‐200c mimics but not other members of the miR‐200 family, significantly repressed the luciferase activity of the reporter gene containing lncRNA‐1604 (Fig. 3B). Moreover, the repressive effects of miR‐200c mimics on lncRNA‐1604 were reversed by introduction of miR‐200c inhibitor (Fig. 3C), and the miR‐200c mimics showed little effect on the mutant lncRNA‐1604 with disrupted seed sequences for miR‐200c (Fig. 3D).
Furthermore, to confirm the direct binding ability of miR‐200c on lncRNA‐1604, we carried out the RIP assays with MS2 binding protein (MS2bp), which specifically bound to RNA containing MS2‐binding sequences (MS2bs). We constructed vectors expressing the lncRNA‐1604 full‐length transcript combined with MS2bs elements (MS2bs‐1604), and cotransfected these vectors with the MS2bp‐YFP vector and miR‐200c mimics into HEK293 cells. The transcript‐specific binding RNA‐protein complexes were immunoprecipitated with YFP antibody, and immunoglobulin G (IgG) was used as a negative control. Results of quantitative PCR demonstrated that miR‐200c was significantly enriched in MS2bs‐1604‐binding RNA compared with the negative control MS2bs‐Renilla luciferase (MS2bs‐RL) RNA (Fig. 3E). Taken together, we found that lncRNA‐1604 functionally interacted with miR‐200c and could potentially serve as a miR‐200c sponge.
It has been reported that the miR‐200 family is involved in various biological processes by targeting ZEB1/2, which are core transcription factors of neural differentiation 20-22. Here, we performed dual‐luciferase reporter gene assays and confirmed the direct repressive effects of miR‐200c on ZEB1/2 (Fig. 3F). Moreover, the repressive effects of miR‐200c on the luciferase activity of ZEB1/2 were rescued by overexpression of lncRNA‐1604 (Fig. 3F). Furthermore, we predicted that lncRNA‐1604 may serve as miR‐200c sponge to regulate ZEB1/2 expression and neural differentiation of mouse ESCs. As expected, knockdown of lncRNA‐1604 (sh1604‐1) (Fig. 3G) and overexpression of miR‐200c (FUW‐miR‐200c) significantly decreased the expression levels of ZEB1/2 (Fig. 3I). Furthermore, immunostaining assays for the expression of ZEB1/2 in mouse teratomas derived from lncRNA‐1604 knockdown (sh1604‐1 and sh1604‐2) and control ESCs (shCtrl) also indicated that knockdown of lncRNA‐1604 significantly decreased the expression level of ZEB1/2 to about 11% for the median intensity (Fig. 3H). Overexpression of lncRNA‐1604 (FUW‐1604) rescued the repressive effects of miR‐200c on ZEB1/2 (Fig. 3I), and introduction of the miR‐200c sponge (sh1604‐1 miR‐200c sponge) reversed the effects of lncRNA‐1604 knockdown (sh1604‐1) on the expression level of ZEB1/2 (Fig. 3J). These findings indicated that lncRNA‐1604 serves as a ceRNA of miR‐200c to regulate the core transcription factors ZEB1/2.
To determine the effects of miR‐200c and ZEB1/2 on the neural differentiation process, we first found that miR‐200c was continuously and highly expressed during neural differentiation of mouse ESCs (Supporting Information Fig. S2A). Then, we stably overexpressed miR‐200c in mouse ESCs (Fig. 3I) and performed neural differentiation with the miR‐200c‐overexpressing (FUW‐miR‐200c) and control (FUW‐Luc) mouse ESCs. Results showed that along with overexpression of miR‐200c, the percentage of Sox1‐GFP cells (Fig. 4A, 4B) was significantly decreased at days 3, 5, and 7 during the neural differentiation process, as similar to that of lncRNA‐1604 knockdown. These findings were confirmed by the immunostaining assay for the protein level of Sox1 and N‐cad in neural progenitor cells generated from miR‐200c‐overexpressing mouse ESCs compared with the control group (FUW‐miR‐200c vs. FUW‐Luc) (Fig. 4C). Quantitative RT‐PCR analysis also revealed that the mRNA levels of neural progenitor cell markers (Sox1, Nestin, Zfp521, Pax6, and N‐cad) were significantly downregulated after overexpression of miR‐200c (FUW‐miR‐200c vs. FUW‐Luc) (Fig. 4D). Collectively, we found that overexpression of miR‐200c severely compromised the neural differentiation process of mouse ESCs, consistent with that of lncRNA‐1604 knockdown, indicating that overexpression of miR‐200c could mimic the repressive effects of lncRNA‐1604 knockdown on neural differentiation.
MiR‐200c and ZEB1/2 play critical role in neural differentiation of mouse embryonic stem cells (ESCs). (A): The representative Sox1‐green fluorescent protein (GFP) images for the effects of miR‐200c overexpression (FUW‐miR‐200c) on the neural differentiation process (days 3, 5, and 7) of mouse ESCs. (B): The fluorescence‐activated cell sorter (FACS) assay of Sox1‐GFP cells for the effects of miR‐200c overexpression on neural differentiation process of mouse ESCs compared with FUW‐Luc. (C): Immunostaining assay for the representative genes Sox1 and N‐cad in neural progenitor cells derived from miR‐200c overexpression (FUW‐miR‐200c) and control (FUW‐Luc) mouse ESCs. (D): Quantitative reverse transcriptase polymerase chain reaction (qRT‐PCR) analysis for the expression level of neural progenitor cell markers (Sox1, Nestin, Zfp521, Pax6, and N‐cad) during the neural differentiation process of mouse ESCs after miR‐200c overexpression compared with FUW‐Luc. (E): qRT‐PCR and Western blotting analysis for the knockdown efficiency of shRNAs targeting ZEB1 (shZEB1‐1 and shZEB1‐2) and ZEB2 (shZEB2‐1 and shZEB2‐2) compared with shCtrl. (F): The FACS assay of Sox1‐GFP cells for the effects of ZEB1 knockdown (shZEB1‐1 and shZEB1‐2) and ZEB2 knockdown (shZEB2‐1 and shZEB2‐2) on the neural differentiation process of mouse ESCs compared with shCtrl. (G): qRT‐PCR analysis for the expression level of neural progenitor cell markers (Sox1, Nestin, Zfp521, Pax6, and N‐cad) during neural differentiation process of mouse ESCs after ZEB1 knockdown (shZEB1‐1 and shZEB1‐2) and ZEB2 knockdown (shZEB2‐1 and shZEB2‐2) compared with shCtrl. (H): Immunostaining assay for the representative genes Sox1 and N‐cad in neural progenitor cells derived from ZEB1 knockdown (shZEB1‐1 and shZEB1‐2), ZEB2 knockdown (shZEB2‐1 and shZEB2‐2), and control (shCtrl) mouse ESCs. Ho.33342 was used for nucleus staining (blue). GAPDH was used as the internal control. Scale bar = 100 μm. Error bars showed SEM (n = 3). *, p < .05; **, p < .01; ***, p < .001.
Furthermore, as ZEB1 and ZEB2 were upregulated during the neural differentiation of mouse ESCs (Supporting Information Fig. S2A), to investigate the exact effects of ZEB1/2, we performed the corresponding neural differentiation experiment with ZEB1 or ZEB2 knockdown mouse ESCs. As shown in Figure 4E, the knockdown efficiency of mRNA and protein levels for the shRNAs targeting ZEB1 (shZEB1‐1 and shZEB1–2) and ZEB2 (shZEB2‐1 and shZEB2‐2), was confirmed by quantitative RT‐PCR analysis and Western blotting compared with the control cells (shCtrl). With these mouse ESCs, we found that knockdown of ZEB1 and ZEB2 efficiently repressed the neural differentiation process, including the percentage of Sox1‐GFP cells (Fig. 4F; Supporting Information Fig. S3) and the expression levels of neural progenitor cell markers (Sox1, Nestin, Zfp521, Pax6, and N‐cad) (Fig. 4G) at days 3, 5, and 7. Consistent with the effects of miR‐200c overexpression and lncRNA‐1604 knockdown on neural differentiation, the percentage of Sox1/N‐cad positive cells of neural progenitor cells derived from ZEB1 or ZEB2 knockdown mouse ESCs (shZEB1‐1, shZEB1–2, shZEB2‐1, and shZEB2‐2) was substantially lower than the control group (shCtrl) (Fig. 4H). Thus, knockdown of ZEB1 or ZEB2 strongly mimicked the repressive effects of miR‐200c overexpression and lncRNA‐1604 knockdown on the neural differentiation process. All these findings indicated that miR‐200c and ZEB1/2 may be the functional downstream targets of lncRNA‐1604 during the neural differentiation process.
To further confirm that miR‐200c and ZEB1/2 are the functional downstream targets of lncRNA‐1604 during the neural differentiation process, we carried out the following analyses with the miR‐200c sponge and ZEB1/2 overexpression in lncRNA‐1604 knockdown mouse ESCs (sh1604‐1). The expression levels of ZEB1 and ZEB2 in the groups of shCtrl FUW‐Luc, sh1604‐1 FUW‐Luc, and sh1604‐1 miR‐200c sponge were observed by Western blotting assays, which confirmed the functional efficiency of the miR‐200c sponge (Fig. 3J). During the neural differentiation process, we found that introduction of the miR‐200c sponge significantly restored the repressive effects of lncRNA‐1604 knockdown on the percentage of Sox1‐GFP cells during neural differentiation of mouse ESCs (Fig. 5A; Supporting Information Fig. S4A). Furthermore, the immunostaining assays of the Sox1/N‐cad positive neural progenitor cells (Fig. 5B) and the expression levels of neural progenitor cell markers (Sox1, Nestin, Zfp521, Pax6, and N‐cad) (Fig. 5C) also confirmed that functional inhibition of miR‐200c by sponge rescued the effects of lncRNA‐1604 knockdown on neural differentiation. These results suggested that lncRNA‐1604 plays a functional role in neural differentiation by competitively binding with miR‐200c.
LncRNA‐1604 orchestrates neural differentiation of mouse embryonic stem cells (ESCs) by targeting miR‐200c/ZEB axis. (A): The fluorescence‐activated cell sorting (FACS) assay of Sox1‐green fluorescent protein (GFP) cells for the effects of the miR‐200c sponge on the neural differentiation process of lncRNA‐1604 knockdown ESCs (sh1604‐1). (B): Immunostaining assay for the representative genes Sox1 and N‐cad in neural progenitor cells derived from lncRNA‐1604 knockdown ESCs with introduction of the miR‐200c sponge (sh1604‐1 miR‐200c sponge), lncRNA‐1604 knockdown ESCs (sh1604‐1 FUW‐Luc), and control ESCs (shCtrl FUW‐Luc). (C): Quantitative reverse transcriptase polymerase chain reaction (RT‐PCR) analysis for the expression level of neural progenitor cell markers (Sox1, Nestin, Zfp521, Pax6, and N‐cad) during the neural differentiation process of lncRNA‐1604 knockdown ESCs (sh1604‐1) after introduction of miR‐200c sponge. (D): The FACS assay of Sox1‐GFP cells for the effects of ZEB1 overexpression (ZEB1), ZEB2 overexpression (ZEB2), or a combination of ZEB1 and ZEB2 (ZEB1 ZEB2) on the neural differentiation process of lncRNA‐1604 knockdown ESCs (sh1604‐1). (E): qRT‐PCR analysis for the expression level of neural progenitor cell markers (Sox1, Nestin, Zfp521, Pax6, and N‐cad) during the neural differentiation process of lncRNA‐1604 knockdown ESCs after ZEB1 overexpression (ZEB1), ZEB2 overexpression (ZEB2), or a combination of ZEB1 and ZEB2 overexpression (ZEB1 ZEB2) compared with “shCtrl FUW‐Luc” (*) or “sh1604‐1 FUW‐Luc” (#). (F): Immunostaining assay for the representative genes Sox1 and N‐cad in neural progenitor cells derived from lncRNA‐1604 knockdown ESCs after ZEB1 overexpression (ZEB1), ZEB2 overexpression (ZEB2), or a combination of ZEB1 and ZEB2 overexpression (ZEB1 ZEB2). GAPDH was used as the internal control. Scale bar = 100 μm. Ho.33342 was used for nucleus staining (blue). Error bars showed SEM (n = 3). The asterisk (*) denotes a significance versus “shCtrl FUW‐Luc,” and the hash mark (#) denotes a significance versus “sh1604‐1 FUW‐Luc”. *, #, p < .05; **, ##, p < .01; ***, ###, p < .001.
As ZEB1 and ZEB2 were downregulated upon knockdown of lncRNA‐1604, we performed functional analysis for the effects of ZEB1 and ZEB2 on lncRNA‐1604 knockdown‐induced neural differentiation repression. The overexpression efficiency of the lentiviral vectors containing ZEB1 (FUW‐ZEB1) and ZEB2 (FUW‐ZEB2) were confirmed by Western blotting (Supporting Information Fig. S4B). During neural differentiation, the decreased effects of lncRNA‐1604 knockdown on Sox1‐GFP cells were significantly restored by overexpression of ZEB1, ZEB2, or especially in combination of ZEB1 and ZEB2 (Fig. 5D; Supporting Information Fig. S4C). Quantitative RT‐PCR analysis of the expression levels of neural progenitor cell markers (Sox1, Nestin, Zfp521, Pax6, and N‐cad) (Fig. 5E) and the immunostaining assays for the Sox1/N‐cad positive neural progenitor cells also confirmed that overexpression of ZEB1 and ZEB2 alleviated the neural differentiation defects triggered by lncRNA‐1604 knockdown (Fig. 5F). Taken together, these findings demonstrated that ZEB1 and ZEB2 act as downstream targets of lncRNA‐1604, and lncRNA‐1604 plays a functional role in neural differentiation by competitive binding with miR‐200c to regulate ZEB1/2.
Accumulating studies have revealed that aberrant neural development is associated with several neurological and psychiatric disorders and results in seriously compromised human health. Direct neural differentiation of ESCs in vitro is becoming a promising strategy to understand neural development and contribute to stem cell therapy 2, 3. Multiple core transcription factors, signaling pathways, and miRNAs have been shown to be required for neural differentiation potential of mouse ESCs 34-37. LncRNAs, as important epigenetic factors for gene expression regulation and cell fate determination, play important role in pluripotency maintenance 11, 12, 19 and multiple lineage differentiation of mouse ESCs 38-40. However, the mechanisms of lncRNAs in neural differentiation of mouse ESCs remain largely unknown. Here, we found that lncRNA‐1604 was primarily expressed in the cytoplasm and highly expressed in mouse ESCs and neural tissues, and further demonstrated that knockdown of lncRNA‐1604 significantly repressed direct differentiation of mouse ESCs into neural progenitor cells, indicating that lncRNA‐1604 is a critical regulator for the early stage of neural differentiation.
Generally, lncRNAs can function both in cis and in trans as decoys, guides and scaffolds by directly interacting with DNA, RNA, or proteins 19, 41-43. In this study, both bioinformatics prediction and experimental verification indicated that lncRNA‐1604 acted as a ceRNA of miR‐200c via a specific binding site. Furthermore, we found that overexpression of miR‐200c could mimic the repressive effects of lncRNA‐1604 knockdown on neural differentiation of mouse ESCs, which firstly demonstrated the function of miR‐200c during neural differentiation of mouse ESCs. Moreover, introduction of the miR‐200c sponge rescued the repressive effect of lncRNA‐1604 knockdown on the transition from mouse ESCs to neural progenitor cells. These findings suggested that lncRNA‐1604 acted as a ceRNA of miR‐200c to orchestrate neural differentiation of mouse ESCs.
It has been indicated that miRNAs play critical role in regulating gene expression at the post‐transcriptional level primarily by directly binding to the 3′UTR of mRNA 17. Among several targets of miR‐200c, zinc finger E‐box‐binding homeobox transcription factor ZEB1 and ZEB2 are the most widely known in human cancers, somatic cell reprogramming, and pluripotency maintenance 20, 22. ZEB1 (also known as Tcf8, δEF1, and Nil‐2α) and ZEB2 (also known as ZFXH1B and SIP1) are two DNA‐binding transcriptional repressors that interact with activated Smads, the transducers of TGF‐β/BMP signaling 44. ZEB2 was implicated in neuroectoderm development in zebrafish 45, xenopus 46, chick 47, and humans 48, and was essential for the development of vagal neural crest precursors and the migratory behavior of the cranial neural crest in mice 49. Here, we found that knockdown of lncRNA‐1604 decreased the expression level of ZEB1 and ZEB2, and it acted as a competitor to restrict miR‐200c activity through the specific binding site to sustain the expression of the core transcription factors ZEB1 and ZEB2. Further analyses showed that knockdown of ZEB1 or ZEB2 significantly repressed neural differentiation of mouse ESCs and mimicked the effects of lncRNA‐1604 knockdown or miR‐200c overexpression. Furthermore, overexpression of ZEB1 and ZEB2 reversed the effect of lncRNA‐1604 knockdown on neural differentiation of mouse ESCs. These findings indicated the critical function of lncRNA‐1604 during neural differentiation was predominantly mediated by the miR‐200c/ZEB axis, providing new evidence for the interaction of lncRNA, miRNA, and transcription factors in the transition from mouse ESCs to neural progenitors.
Recently, it has been reported that miR‐200c repress neural differentiation process of human ESCs through targeting ZEB1 and ZEB2 23. Here, we confirmed the critical function of the miR‐200c/ZEB axis during neural differentiation process of mouse ESCs. However, the expression level of miR‐200c was observed to be continuously high during mouse neural differentiation, which was different from human neural differentiation 23, 50. In this study, we found that the expression pattern of miR‐200c was similar to that of lncRNA‐1604, which sequestering miR‐200c to ensure the normal up‐regulation of ZEB1/2 was at least one important mechanism of lncRNA‐1604 during neural differentiation of mouse ESCs. As there was no homolog of lncRNA‐1604 in human, this study indicated the critical function and regulatory mechanism of a specific lncRNA‐1604/miR‐200c/ZEB axis during neural differentiation of mouse ESCs. Additionally, lncRNA‐1604 may be also required for other functions such as to re‐direct the binding of other neural TFs.
In addition, to assess the critical function of the lncRNA‐1604/miR‐200c/ZEB axis during neural differentiation of mouse ESCs, we performed a further exploration to analyze the correlation of this key axis with neurological and psychiatric disorders. Interestingly, from the differentially expressed genes and lncRNAs of Huntington's disease, we found that the expression level of lncRNA‐1604 (also termed predicted gene GM11627) and ZEB1/2 significantly decreased in the cortex and striatum of 6‐ to 10‐month‐old mice with Huntington's disease compared with the normal group 51. It has also been reported that the expression level of miR‐200c was increased in the cortex and striatum of mice with Huntington's disease compared with normal group 52. These findings provide new evidence for the critical function of the lncRNA‐1604/miR‐200c/ZEB axis during neural differentiation and suggest that lncRNA‐1604/miR‐200c/ZEB dysregulation may be related to abnormal neural differentiation and neurodegenerative diseases, such as Huntington's disease. Further exploration of the function of the lncRNA‐1604/miR‐200c/ZEB axis in neural differentiation and development may systematically uncover the critical function and mechanism of lncRNA‐1604 in vivo.
We found that lncRNA‐1604, a novel cytosolic expressed lncRNA, was required for the neural differentiation process of mouse ESCs. It regulates the core transcription factors ZEB1 and ZEB2 by sequestering miR‐200c in the neural differentiation process. These findings not only indicate that lncRNA‐1604 orchestrates neural differentiation of mouse ESCs through the miR‐200c/ZEB axis but also provide new evidence for the regulatory mechanisms of lncRNAs in development and diseases associated with the nervous system.
This work was supported by grants obtained from the Ministry of Science and Technology of the People's Republic of China (2016YFA0101300), the National Natural Science Foundation of China (31371510, 31571519, 31471250, 81530042, 31210103905, 31571529, and 31571390), the Science and Technology Commission of Shanghai Municipality (15JC1403201), and the Fundamental Research Funds for the Central Universities (2000219136 and 1500219106).
R.W. and C.L.: conception and design, collection and assembly of data, data analysis and interpretation, manuscript writing; X.L., G.L., Y.L., and J.Q.: collection of data, data interpretation, manuscript writing; M.B., Z.W., X.G., D.Y., Z.J., Y.Y., and C.X.: data interpretation, manuscript writing; G.W. and J.K.: conception and design, financial support, provision of study material, assembly of data, data analysis and interpretation, manuscript writing, final approval of manuscript.
The authors indicated no potential conflicts of interest.
Additional Supporting Information may be found in the online version of this article.
| Filename | Description |
|---|---|
| stem2749-sup-0001-suppinfo.docx15.4 KB | Supporting Information |
| stem2749-sup-0002-suppinfoFigS1.tif3.3 MB | Supporting Information Figure S1 |
| stem2749-sup-0003-suppinfoFigS2.tif7.4 MB | Supporting Information Figure S2 |
| stem2749-sup-0004-suppinfoFigS3.tif26.2 MB | Supporting Information Figure S3 |
| stem2749-sup-0005-suppinfoFigS4.tif13.5 MB | Supporting Information Figure S4 |
| stem2749-sup-0006-suppinfoTabS1.docx13.1 KB | Supporting Information Table S1 |
| stem2749-sup-0007-suppinfoTabS2.docx14.4 KB | Supporting Information Table S2 |
| stem2749-sup-0008-suppinfoTabS3.xlsx44.2 KB | Supporting Information Table S3 |
Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
联系客服
