Stem Cell Research (2013) 11, 1299–1313

Available online at www.sciencedirect.com

ScienceDirect

www.elsevier.com/locate/scr

MicroRNA-137 represses and Tbx3 during differentiation of mouse embryonic stem cells Ke Jiang a, Chunyan Ren b, Venugopalan D. Nair a,⁎ a Department of Neurology, Mount Sinai School of Medicine, New York, NY 10029, USA b Graduate School of Biological Sciences, Mount Sinai School of Medicine, New York, NY 10029, USA

Received 19 December 2012; received in revised form 5 September 2013; accepted 5 September 2013 Available online 13 September 2013

Abstract MicroRNA-137 (miR-137) has been shown to play an important role in the differentiation of neural stem cells. Embryonic stem (ES) cells have the potential to differentiate into different cell types including neurons; however, the contribution of miR-137 in the maintenance and differentiation of ES cells remains unknown.Here,weshowthatmiR-137 is mainly expressed in ES cells at the mitotic phase of the cell cycle and highly upregulated during differentiation. We identify that ES cell transcription factors, Klf4 and Tbx3, are downstream targets of miR-137, and we show that endogenous miR-137 represses the 3′ untranslated regions of Klf4 and Tbx3. Transfection of ES cells with mature miR-137 RNA duplexes led to a significant reduction in cell proliferation and the expression of Klf4, Tbx3, and other self-renewal . Furthermore, we demonstrate that increased miR-137 expression accelerates differentiation of ES cells in vitro. Loss of miR-137 during ES cell differentiation significantly impeded neuronal expression and morphogenesis. Taken together, our results suggest that miR-137 regulates ES cell proliferation and differentiation by repressing the expression of downstream targets, including Klf4 and Tbx3. © 2013 Elsevier B.V. All rights reserved.

Introduction et al., 2012; Pauli et al., 2011). The self-regulatory networks formed by transcription factors such as Oct4, , Nanog, Klf4, Embryonic stem (ES) cells have the remarkable ability Esrrb, Tbx3, and Tcf3 regulate a wide range of downstream to replicate indefinitely while retaining the capacity to genes required for self-renewal and pluripotency (Boyer et al., differentiate into functionally distinct cell types including 2005; Chambers and Tomlinson, 2009; Ivanova et al., 2006; neurons. The pluripotency and self-renewal of ES cells Kim et al., 2008; Silva and Smith, 2008). In addition to the are regulated by a core transcriptional circuitry comprising transcription factors, epigenetic changes associated with of transcription factors, chromatin modulators, and small posttranslational modifications of histone proteins are known non-coding microRNAs (miRNAs) (Dejosez and Zwaka, 2012; Li to regulate in ES cells (Li et al., 2012).

Abbreviations: ES cells, embryonic stem cells; NC, neuronally differentiated ES cells; EB, embryoid body; LIF, leukemia inhibitory factor; miRNA, microRNA; 3′UTR, 3′ untranslated region; LSD1, lysine specific demethylase 1; Tuj1, neuronal class III β-tubulin 3; RT, reverse transcription; qPCR, quantitative real-time polymerase chain reaction; ddPCR, droplet digital PCR; DAPI, 4,6-diamidino-2-phenylindole dihydrochloride; PBS, phosphate buffered saline; SSC, saline-sodium citrate. ⁎ Corresponding author at: Dept. of Neurology, Mount Sinai School of Medicine, Annenberg 14-72, Box 1137, One Gustave L. Levy Place, New York, NY 10029, USA. Fax: +1 212 289 4107. E-mail address: [email protected] (V.D. Nair).

1873-5061/$ - see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.scr.2013.09.001 1300 K. Jiang et al.

Moreover, by directly interacting with ES cell transcription Materials and methods factors, miRNAs are shown to regulate the central signaling circuitry in ES cells (Pauli et al., 2011). The genetic ES cell culture and differentiation deletion of key miRNA processing enzymes such as Dicer or Dgcr8 has been shown to impair cell cycle progression and − − R1 ES (ATCC) and Dgcr8 / (Novus Biologicals) cells were cause defective differentiation (Murchison et al., 2005; maintained on mitomycin C treated MEF feeder layers and Wang et al., 2008). Studies on the role of miRNAs in ES cells differentiated in vitro as described previously (Melton et suggest that the regulatory function of the miRNA is often al., 2010; Nair, 2006). Briefly, undifferentiated ES cells regulated through, and controlled by, various transcrip- were grown on gelatin-coated tissue culture plates in the tion factors important for self-renewal and pluripotency presence of 1000 U/ml leukemia inhibitory factor (LIF; (Marson et al., 2008; Melton et al., 2010; Tay et al., 2008; Chemicon) in ES cell medium. To induce embryoid body Xu et al., 2009). An understanding of the interactions (EB) formation, the cells were trypsinized and dissociated among these factors will be critical for the development of to a single-cell suspension and plated in ES cell medium improved strategies to reprogram differentiated cells or without LIF onto non-adherent bacterial culture dishes at a induce differentiation of pluripotent ES cells to function- density of 2–2.5 × 106 cells/dish. Four-day-old EBs were ally distinct cell types. plated onto adherent tissue culture dishes and maintained miRNAs are short non-coding RNAs that can modulate in ES cell medium for 24 h. The selection of nestin-positive gene expression by binding to the complementary se- cells was initiated by replacing the ES cell medium with quences on target sites in the 3′ untranslated regions serum-free insulin/transferrin/selenium medium. Cells grown (UTRs) of messenger RNA (Bartel, 2004). There is strong in differentiation medium for 12–14 days were used in all evidence that miRNAs play an important role in the experiments. maintenance of ES cells (Pauli et al., 2011). In ES cells, a majority of the miRNA species are produced by 4 miRNA loci (the miR-21, the miR-17-92 cluster, the miR-15b-16 cluster and the miR-290-295 cluster) that regulate cell Western immunoblotting cycle progression and oncogenesis (Calabrese et al., 2007). During differentiation, a number of miRNAs exhibit distinct Total cell extracts were prepared and Western immunoblotting expression patterns and have been shown to fine tune or wascarriedoutasdescribedinourpreviouspublication(Nair et α restrict cellular identities by targeting important tran- al., 2012). Western blots of Klf4, Tbx3, LSD1, CoREST, RXR , β scription factors or key pathways (Stefani and Slack, 2008). GAPDH, and -actin were performed using anti-Klf4 (Millipore; For example, miRNA-145 has been shown to repress Oct4, 09-821), anti-Tbx3 (SCB: sc-48781), anti-LSD1 (CST; 2139), anti- α Sox2, and Klf4 during differentiation of human ES cells CoREST (Millipore; 07-455), anti-RXR (SCB; sc-553), anti- β into mesoderm and ectoderm lineages (Xu et al., 2009). GAPDH (SCB; sc-166545), and -actin (SCB; sc-166545) Furthermore, miR-134, miR-296, and miR-470 target the antibodies. All horseradish peroxide-conjugated secondary transcription factors Nanog, Oct4, and Sox2 to promote antibodies were from Santa Cruz Biotechnology. retinoic-acid-induced differentiation of mouse ES cells (Tay et al., 2008). The contribution of additional miRNAs involved in the direct repression of ES cell transcription factors during Quantitative real-time PCR neuronal differentiation remains unknown. We searched for miRNAs that show increased expression levels during neuronal Quantitative real-time PCR (RT-qPCR) was carried out to differentiation and are predicted to target ES cell genes. The measuregeneexpressionasdescribedinourpreviouspublica- expression of miR-137 has been found to increase during tion (Nair et al., 2012). Briefly, RNA was isolated using an differentiation of neural stem cells into neurons (Sun et al., Absolutely RNA preparation kit(AgilentTechnologies)and 2011b; Szulwach et al., 2010). Moreover, miR-137 is one of reverse transcribed using the Superscript™ III first-strand the two miRNAs in ES cells that are co-occupied by the key synthesis system (Invitrogen). SYBR green qPCR reactions on transcription factors: Oct4, Sox2, and Nanog (Boyer et al., resulting cDNAs were performed on an ABI 7900HT (Applied 2005). These findings suggest that miR-137 may be an important Biosystems). Gene-specific primers used for RT-qPCR are shown component of the transcriptional regulatory circuitry, and that in Table 1. β-Actin, α-tubulin, and Rps11 were used as internal temporal expression of miR-137 may influence proliferation and normalization controls. differentiation of ES cells. For RT-qPCR assays of miRNAs, total RNA was prepared Here, we show that miR-137 is mainly expressed in ES using mirVANA miRNA isolation kit (Ambion). cDNA was cells at mitotic stages of the cell cycle and is significantly synthesized from 100 ng of total RNA using miRNA specific upregulated when ES cells are differentiated into neuronal RT primers and the expression was quantified using TaqMan lineage. We demonstrated that miR-137 directly targets the MicroRNA Kits according to the manufacturer's protocol ES cell transcription factors, Klf4 and Tbx3, as well as a (Applied Biosystems). RT-qPCR assays using TaqMan Assay known miR-137 target, lysine specific demethylase 1 (LSD1) Kits, mmu-miR137 (ID 001129), hsa-miR16 (000396), and (Balaguer et al., 2010; Sun et al., 2011b). The repressive mouse snoRNA (001232) were used to determine the effect of miR-137 is mediated by its binding to the target expression of miR-137, miR-16, and snoRNA, respectively. sites present in the 3′ untranslated regions (3′UTRs) of the Relative fold mRNA and miRNA levels were determined by mRNAs. Furthermore, we found that miR-137 disrupts ES the 2−ΔΔCT (cycle threshold) method (Nair et al., 2012). cell self-renewal and accelerates differentiation of ES cells The expression levels of each miRNA in undifferentiated ES in vitro. cells (ESC) were arbitrarily set to one. MicroRNA-epresses Klf4 and Tbx3 during differentiation of mouse embryonic stem cells 1301

Table 1 RT-qPCR primer sequences. Gene name Sense primer Antisense primer Klf4 GTCAAGTTCCCAGCAAGTCAG CATCCAGTATCAGACCCCATC Tbx3 TGGCTCAGTGTCCTTGTCAC ACTGGAATGGAGAGACCTTGG LSD1 CACAGCAGTCCCCAAGTATGT GCCTCTGCTGTCAAACTAGGA Oct 4 CCCACTTCACCACACTCTACT GCTCCAGGTTCTCTTGTCTA Nanog CCACCAGGTGAAATATGAGAC GGCTCACAACCATACGTAAC Sox2 CCGTGATGCCGACTAGAAAAC AGCGCCTAACGTACCACTAG REST AAGGAACCACCTCCCAGTATG CGTCTGAGCAACCTCAATCTG CoREST GCCTCATTCACCTTTCCTTC CCAACCCTTCTTTCTCCAAAC Calb1 TGCCAGAGAGTATGACCATAG GGAAGTAGGCAAACTAGACAC Kcnn1 TCACCTTCAACACACGCTTC GCTGGTCACTTCCTGCTTATC RXR GCAAACACAAGTACCCTGAGC CATCTCCATGAGGAAGGTGTC Scn3a TTGCCTAACATGATACAACTC TTGGCTCAATAACGAAGTAG Slc1a3 TGTGCTTGTTTATGTCCCTACC TCTCCTGCTGTGTTTTCTTCC NeuroG2 GACTCCACTTCACAGAGCAGAG CACGCCATAGTCCTCTTTGAC Rps11 GCACATTGAATCGCACAGTC CGTGACGAAGATGAAGATGC α-Actin CTGCGCAAGTTAGGTTTTGTCAA GCTGCCTCAACACCTCAAC β-Tubulin AG CTGGAGCAGTTTGACGACAC TGCCTTTGTGCACTGGTATG TTACACACCACTGACGCACAC AGGCTATGAGGAGGCTTTGG Sox17 GCAACTACCCCGACATTTGAC CAAAACAGCTTGACGCTAAGG Mixl1 TCTTCCGACAGACCATGTACC TGAAATGACTTCCCACTCTGG Sox1 TCTCCAACTCTCAGGGCTACA ACTTGACCAGAGATCCGAGG Tuj1 CCACCTTCATTGGCAACAGC GCTTCCGATTCCTCGTCATCA

Droplet digital PCR GCAG-3′, according to the manufacturer's instructions (Promega). Klf4 3′UTR luciferase reporter was constructed miRNA levels were measured by digital PCR using the by cloning the PCR-amplified DNA corresponding to 714– QuantaLife droplet digital PCR (ddPCR) system (Nair et al., 790 bp of mouse Klf4 3′UTR, flanking the miR-137 binding 2012). Briefly, 20 μl of ddPCR mixtures was prepared by site (753–759 bp) into the pDrive vector. The Klf4 3′UTR combining the respective reverse transcription (RT) product, was then excised by HindIII/Spe1 digestion and cloned primers, and probes with the QuantaLife master mix. Approx- into the pMIR-REPORT luciferase vector. We deleted the imately 10,000 to 14,000 monodispersed droplets for each miR-137 target site 5′-GCAATAA-3′ at the KLf4 3′UTR by sample were prepared using the QuantaLife droplet generator. using a site-directed mutagenesis kit with overlapping The droplets were transferred to a 96-well PCR plate and mutated primers 5′-TTATGCACTGTGGTTTCAGATGTTTTGT amplified to endpoint using the following conditions: 95 °C for ACAAT-GGTTTATT-3′ and 5′-AATACGTGACACCAAAGTCTAC 10 min, 40 cycles of 94 °C for 30 s, 60 °C for 1 min, and 98 °C AAAACATGTTACCAAATAA-3′. Nucleotide sequences of all for 10 min in a standard thermal cycler (PTC100; MJ Research). 3′UTR constructs were confirmed by DNA sequencing. The Tbx3 Plates were read in a QuantaLife droplet reader, and the 3′UTR wild type with miR-137 binding site (8 bp GCAATAAC, concentrations (copy numbers) of the targets in the samples 862–869) and Tbx3 3′UTR with mutant miR-137 sites (866– were determined using QuantaSoft software. The expression 869 bp), Oct4 3′UTR, and Sox2 3′UTR constructs used in the levels of each miRNA in vehicle-treated cells (control) were study were reported previously (Xu et al., 2009; Zhang et al., arbitrarily set to one. 2011).

3′UTR luciferase reporter constructs Luciferase reporter assay

TomaketheluciferasereporterconstructswithLSD13′ ES cells growing on gelatinized plates in LIF media were used UTR, full-length cDNA, corresponding to the 3′UTR of LSD1 in transfection experiments. We plated 30,000 wild type (2665–2999 bp), was amplified from mouse ES cell mRNA (R1 ES) or 40,000 Dgcr8−/− ES cells per well in a 48-well plate by PCR. The product was sub-cloned into the pDrive vector and let them grow for at least 16 h before transfection. (Qiagen), and we confirmed the sequences. The 3′UTR For transfection experiments, we transfected a mixture of was then excised by Mlu1-HindIII digestion and cloned into 250 ng of reporter to each well using Lipofectamine 2000, the pMIR-REPORT Luciferase vector (Ambion). To generate according to the manufacturer's instructions (Invitrogen). aLSD13′UTR luciferase reporter construct-lacking miR- The pSV-β-Galactosidase Vector (Promega) was used as a 137 binding site, we deleted the miR-137 target site control plasmid for monitoring transfection efficiencies. In 5′-TAGCAATA-3′ by using a site-directed mutagenesis kit co-transfection experiments, luciferase reporters were with overlapping mutated primers 5′-CTGCCACGTAAG transfected with 10–200 nM of mature miR-137 RNA duplexes CAAGCTCTTCCTAGATCCTACTGAGAAACTCCATG-3′ and 5′- (UUAUUGCUUAAGAAUACGCGUAG; Dharmacon, C-310413-05), CATGGAGTTTCTCAGTAGGATCTAGGAAGAGCTTGCTTACGTG miR-137 inhibitor (Dharmacon, IH-310413-07), or a combination 1302 K. Jiang et al. of both using Lipofectamine 2000. Cells were lysed 24–48 h state were confirmed as described previously (Nair et al., later, and processed for luciferase and β-galactosidase assays 2012) and used for the analysis of miR-137 expression. using Dual-Luciferase Reporter Assay System (Promega) and β -Galactosidase Enzyme Assay System (Promega), respectively. Cell proliferation assay The luciferase activity was measured using a Mithras LB 940 luminometer (Berthold Technologies), and β-galactosidase For cell proliferation analysis, cells were plated at a density of activity was measured at 420 nm using a spectrophotometer 2×104 cells/well on 6, 96-microwell cell-culture plates (in (Spectra Max Gemini XS, Molecular Devices). 100 μl of medium). The next day (D0), cells were transfected with mature miR-137 RNA duplexes (100 nM), miR-137 inhibitor Fluorescent in situ hybridization (50 nM), or a microRNA inhibitor negative control (Dharmacon, IN-001005-01; 50 nM) and grown for 5 days. Media was changed Mouse ES cells were grown on gelatin-coated cover glasses daily and the cell proliferation was determined on day 0, 1, 2, which were fixed and permeabilized as described previ- 3, and 4 using CellTiter-Blue™ reagent (Promega, Madison, WI). ously (Nair et al., 2012). For the detection of miR-137, the The fluorescence was quantified using a spectrofluorometer cellswerehybridizedwith5′-fluorescein labeled LNA (Spectra Max Gemini XS, Molecular Devices) by measuring the detection probe (Exiquon, #38510-04) overnight in 50% fluorescence at excitation/emission wavelengths of 570/ formamide/5× SSC at 50 °C. The cells were washed three 590 nm. Data are expressed as a percentage of the times with 50% formamide/2× saline-sodium citrate (SSC), vehicle-treated controls, and values represent the three times with 2× SSC, and three times 0.2× SSC at means ±s.e.m. from eight microwells from each of two 50 °C. The cells were washed with phosphate buffered independent experiments (n = 16). saline containing 0.1% Tween 20 (PBST), and the nuclei were μ stained with 1 g/ml (in PBS) of the fluorescent DNA dye 4, Statistical analysis 6-diamidino-2-phenylindole dihydrochloride (DAPI) for 10 min and mounted in the Vectashield (Vector Laboratories) mount- The data were analyzed using the GraphPad Prism data ing medium. The slides were examined under an Olympus analysis program (GraphPad Software, San Diego, CA). For (BX65) upright epifluorescence microscope. the comparison of statistical significance between two groups, Student's t tests for paired and unpaired data were Alkaline phosphatase staining used. For multiple comparisons, one-way ANOVA followed by post hoc comparisons of the group means according to the R1 ES cells were transfected with mature miR-137 duplexes, method of Tukey was used. inhibitor, or a combination of both, and grown in complete ES medium and LIF for 5 days. Staining for alkaline phosphatase in these cells was carried out using an alkaline phosphatase Results detection kit according to the manufacturer's instructions (Chemicon, Millipore). miR-137 is enriched in ES cells at the mitotic phase of the cell cycle Immunocytochemistry miRNAs are small non-coding RNAs that are shown to regulate pluripotency and self-renewal by interacting with the 3′UTR of Cells were permeabilized and incubated with primary anti- key transcription factors in ES cells (Pauli et al., 2011). To bodies at 4 °C overnight as described previously (Nair et al., assess the potential miRNA interaction with key ES cell genes, 2012). Primary antibodies used were a rabbit polyclonal we sought miRNAs that are regulated by ES cell transcription anti-LSD1 (dilution 1:400), anti-Tuj1 mouse monoclonal anti- factors and could interact with RNA recognition sites in the 3′ body (Covance; MMS-435P; 1:1000), anti-Nanog (BD Biosci- UTR of these genes. We performed bioinformatics analyses ence; 560259; 1:400), anti-histone H3 (CST; 9715; 1:500), and using TargetScan (Lewis et al., 2003) to identify potential anti-cytochrome c monoclonal antibody (BD Pharmingen; miRNA candidates that have the capacity to interact with the 3′ 556432, 1:400). After washing with PBST (PBS with 0.1% UTR of ES cell genes. One candidate miRNA, miR-137, was Tween20), the cells were incubated at room temperature for selected for further analysis on the basis of its regulation by 2 h with Alexa Fluor 568 donkey anti-rabbit IgG or Alexa Fluor Oct4, Sox2, and Nanog (Boyer et al., 2005)andthepresenceof 488 goat anti-rabbit/mouse IgG (Molecular Probes, dilution target sites in the 3′UTR of many pluripotency and self-renewal 1:400) as indicated. The slides were processed and images genes (Ramalho-Santos et al., 2002; Wang et al., 2008), such as were acquired as described in our previous publication (Nair et Klf4, Klf2, Tbx3, Akt2, GSK-3β, Cdk6, Cdc42, and CCNY mRNA. al., 2012). TaqMan reverse transcription-PCR, droplet digital PCR, and in situ hybridization were used to analyze miR-137 expression in Cell cycle synchronization mouse ES cells. We also amplified miR-16 and snoRNA in these samples. We found that miR-137 is expressed at relatively To synchronize cells to metaphase, R1 ES cells were first low levels in undifferentiated R1 ES cells (Fig. 1A). In situ grown to about 60% confluency in complete ES medium. hybridization of miR-137, coupled with analysis of the nuclear Prior to the addition of 100 ng/ml nocodazole or vincristine, morphology characteristics of the mitotic phase in ES cells the medium was replaced with serum-free ES cell medium growing under self-renewal conditions, demonstrated that and incubated for 16–18 h. Synchronization and cell cycle miR-137 is highly expressed in cells at all stages of mitosis and MicroRNA-epresses Klf4 and Tbx3 during differentiation of mouse embryonic stem cells 1303

Figure 1 miR-137 is highly expressed in M phase ES cells. (A) The relative levels of miR-137 increased after the differentiation of ES cells. In TaqMan real-time RT-qPCR, miR-137, miR-16, and snoRNA levels were quantified in undifferentiated (ESC) and neuronally differentiated R1 ES cells (NC). The expression levels of each miRNA in undifferentiated ES cells (ESC) were arbitrarily set to one. The error bar represents the standard deviation (SD) from three independent experiments. **, P b 0.001. (B) miR-137 is mainly expressed in M phase cells. Expression of miR-137 in R1 ES cells growing under self-renewal condition was assessed by in situ hybridization using fluorescein-labeled miR-137 detection probe (green), while the blue fluorescent signal was from nuclear DNA counterstained with DAPI. Scale bar, 5 μm. (C) Increased expression of miR-137 in cells synchronized at metaphase by nocodazole and vincristine treatment (100 ng/ml; 16 h) was confirmed by droplet digital PCR (ddPCR). Data are the means ± SD (n = 4) from two independent experiments. *, P b 0.01.

nearly undetectable in G1, S, and G2 phase cells (Fig. 1B). We differentiation protocol that includes EB formation, selection, found that all cells showing mitotic nuclear morphology, as expansion of nestin positive cells, and differentiation into described in our recent report (Nair et al., 2012), showed neuronal cells (Nair, 2006). Consistent with reported studies, increased expression of miR-137. These results suggest that we observed a significant increase (~27 fold) in miR-137 miR-137 is expressed in a cell cycle-dependent manner in ES expression upon differentiation of ES cells into neural cells cells. (NC) (Fig. 1A). However, no increase in the levels of miR-16 was To further confirm the expression of miR-137 in mitotic observed between ES cells and neuronally differentiated cells cells, we synchronized ES cells to G2/M phases of the cell (Fig. 1A). These findings suggest that the proper temporal cycleandtheexpressionofmiR-137wasdeterminedbytwo expression of miR-137 may influence ES cell proliferation and independent methods such as qPCR and droplet digital dPCR differentiation. (ddPCR). As shown in Fig. 1C, cells treated with nocodazole and vincristine showed significantly higher miR-137 expression levels. However, no changes in the levels of miR-16, a microRNA miR-137 represses luciferase reporters containing highly expressed in ES cells were observed in cells synchronized Klf4 and Tbx3 3′UTRs to G2/M phase (Fig. 1C). The localization of miR-137 in mitotic cells by in situ hybridization (Fig. 1B) and the increased levels of miRNAs bind to partially complementary target sites present miR-137 in mitotically synchronized cells (Fig. 1C) suggest that in the 3′UTRs of messenger RNA, which results in either miR-137 is expressed in a cell cycle-dependent manner in ES degradation of the target mRNAs ortranslationalrepression of cells. the encoded proteins (Bartel, 2004). By bioinformatics analysis miR-137 has been shown to inhibit proliferation and induce using TargetScan, we found 971 conserved miR-137 targets in differentiation of tumor cells (Kozaki et al., 2008; Silber et al., the mouse genome and identified many potential candidate 2008). miR-137 is a known regulator of neuronal development mRNAs that have the capacity to directly interact with miR-137 and its expression has been found to increase during neuronal inEScells.Twocandidategenes,Klf4andTbx3,wereselected differentiation (Ripke et al., 2011; Silber et al., 2008; Smrt et for further analysis on the basis of their role in ES cell al., 2010; Sun et al., 2011a). We used a stage-controlled ES cell self-renewal and pluripotency and the presence of target 1304 K. Jiang et al.

“seed sequences” in the 3′UTR of these mRNAs (Fig. 2A). We cells, we constructed luciferase reportersthathaveeitherthe also used LSD1, a known miR-137 target (Balaguer et al., 2010; wild-type 3′UTR of these genes or the mutant 3′UTR with Sun et al., 2011b), as a control. To investigate whether Klf4, altered or deleted miR-137 target sites (Fig. 2A).Inorderto Tbx3, and LSD1 can be directly targeted by miR-137 in ES evaluatethespecificfunctionofmiR-137,weusedDgcr8−/−

Figure 2 (A) The miR-137 target site in the Klf4, Tbx3, and LSD1 3′UTR as predicted by TargetScan. The underlined nucleotides (target sites) were deleted in the mutant 3′UTR constructs. (B–G) miR-137 represses Klf4, Tbx3, and LSD1 3′UTR in the luciferase reporter assay. Dgcr8−/− ES cells were co-transfected with indicated wild type 3′UTR reporters or constructs with deleted miR-137 binding sites (3′UTRΔ), and mature miR-137 RNA duplexes, miR-137 inhibitor, or a combination of both. Luciferase activity was determined at 24 h after transfection and represented as the relative level to cells transfected with the basal luciferase reporter without UTR (control). (H) Oct4 3′ UTR luciferase construct was used as a miR-137 non-target control. The error bar represents the SD of three independent experiments (n =6).Student'st test compared two data sets marked by brackets in the panel. *, P b 0.001. (I–N) Endogenous miR-137 represses the 3′UTRs of Klf4, Tbx3, and LSD1 in ES cells. R1 ES cells were transfected with indicated 3′UTR reporter constructs and treated with nocodazole (100 ng/ml; 16 h) after 24 h of transfection. The difference between 3′UTR mutant and wild-type luciferase reporters of Klf4 (I and J), Tbx3 (K and L), and LSD1 (M and N) reflects the magnitude of repression. The error bar represents the SD from three independent experiments. Student's t test compared two data sets marked by brackets in the panel. *, P b 0.01. MicroRNA-epresses Klf4 and Tbx3 during differentiation of mouse embryonic stem cells 1305 mouse ES cells because these cells are relatively deficient of which lacks a miR-137 target site were not repressed by endogenous miRNAs (Melton et al., 2010). Dgcr8−/− ES cells miR-137. These results suggest that a target site is required for were co-transfected with luciferase constructs and mature the repression of the reporters by miR-137 (Figs. 2C, E, G, and miR-137 RNA duplexes, small, double-stranded RNA molecules, H). Taken together, these results show that miR-137 directly designed to mimic endogenous mature miR-137 functions when targets Klf4, Tbx3, and LSD1 3′UTRs in ES cells. transfected into cells. A luciferase reporter with no 3′UTR, We then studied the role of endogenous miR-137 in which is resistant to any miRNA repression in ES cells (negative repressing the Klf4, Tbx3 and LSD1 3′UTR reporters in wild control), and transfection without mature miR-137 RNA type ES cells (R1) under self-renewal conditions (Nair et al., duplexes (mock transfection) were used as controls. The 2012). At 48 h after transfection, there was a small but optimum concentration of mature miR-137 RNA duplexes significant repression of the wild-type 3′UTR luciferase required for the repression of the reporter constructs was reporter activities of Klf4, Tbx3, and LSD1 in comparison to determined by transfecting different concentrations ranging the no-UTR control (Figs. 2I, K, and M). In contrast, the from 10 to 200 nM with LSD1 3′UTR plasmid. We selected respective mutant reporters with deletion of the miR-137 100 nM of mature miR-137 RNA duplexes for further studies target site had significantly less repression than the wild basedonthemaximumrepressionobservedwithwildtype type reporters (Figs. 2J, L, and N). We further addressed the luciferase reporter activity without affecting the control and dependence of 3′UTR reporter repression with the level of mutant reporter expression. Transfection of miR-137 signifi- miR-137. The expression of miR-137 has been shown to be cantly reduced the luciferase activity of the wild-type reporter upregulated by treatment with nocodazole (Fig. 1B). At 24 h constructs of Klf4, Tbx3, and LSD1 (Figs. 2B, D, and F). No after transfection with 3′UTR constructs, the cells were reduced luciferase activity was observed when the constructs synchronized to M phase by treatment with nocodazole were cotransfected with miR-137 inhibitor, a chemically as described previously (Nair et al., 2012). Nocodazole modified, single stranded nucleic acid that was designed to treatment significantly increased the repression of 3′UTR specifically bind to and inhibit the activity of miR-137 (Figs. 2B, constructs compared to non-synchronized cells (Figs. 2I, K, D, and F). Moreover, cotransfection of mature miR-137 RNA and M). However, no significant repression was observed duplexes with miR-137 inhibitor completely abolished the with mutant 3′UTR reporters (Figs. 2J, L, and N). The repression mediated by miR-137 (Figs. 2B, D, and F). Mutant correlation between the increased levels of miR-137 and the reporters and an Oct4 3′UTR luciferase reporter construct repression of the wild type target 3′UTR reporters without

Figure 3 miR-137 represses endogenous Klf4, Tbx3, and LSD1 mRNA and protein. (A–C) Relative mRNA levels of Klf4 (A), Tbx3 (B), and LSD1 (C) in Dgcr8−/− ES cells transfected with mature miR-137 RNA duplexes, miR-137 inhibitor, or the combination of both and mRNA levels were determined by RT-qPCR after 48 h of transfection. (D) The Klf4, Tbx3, and LSD1 protein levels were determined by Western blotting following 48 h after transfection with mature miR-137 RNA duplexes. Lane 1, control; lane 2, miR-137; lane 3, miR-137 inhibitor; lane 4, combination of both transfected cells. β-Actin was used as loading control. Reduction of Klf4, Tbx3, and LSD1 mRNA levels in miR-137 transfected cells correlates with the reduction of the corresponding proteins in these cells. (E) The Oct4 mRNA level measured in these cells remained unchanged in any of the treatment groups. (F–G) Expression of Nanog (F) and Sox2 (G) were significantly repressed in cells transfected with miR-137. All real-time PCR experiments were performed three times in triplicate. Student's t test compared two data sets marked by brackets in the panel. *, P b 0.001. (H) miR-137 had no effect on the expression of Sox2 3′ UTR luciferase construct. The error bar represents the SD of three independent experiments (n = 6). 1306 K. Jiang et al. affecting the expression of mutant reporters in nocodazole differentiation of cells transfected with miR-137 and not by treated cells suggests that endogenous miR-137 also targets cell death. To rule out the non-specific effects of mature the 3′UTRs of Klf4, Tbx3, and LSD1. miR-137 RNA duplexes, we co-transfected duplexes with miR-137 inhibitor, this completely prevented the loss of Overexpression of miR-137 reduces the levels of alkaline phosphatase staining caused by miR-137 (Figs. 4B endogenous Klf4 and Tbx3 in ES cells and C). We also stained these cells for the self-renewal marker, Nanog; LSD1, a direct target of miR-137; and cytochromec,asacontrol.AsshowninFig. 5, transfection miRNAs can downregulate gene expression by one of two of mature miR-137 duplexes significantly reduced the levels post-transcriptional mechanisms: mRNA cleavage or transla- of Nanog and LSD1 (Fig. 5A) without affecting the levels of tional repression (Bartel, 2004). Because the luciferase assay cytochrome c in these cells (Fig. 5B). Taken together, our does not readily distinguish between these mechanisms, we results suggest that temporal expression of miR-137 plays further examined how miR-137 represses the endogenous animportantroleinthenormalcellcycleprogressionofES Klf4, Tbx3, and LSD1 in ES cells. The Dgcr8−/− ES cells were cells, and its continued expression most likely inhibits cell transfected with mature miR-137 duplexes, and we mea- proliferation. sured the mRNA and protein levels of endogenous Klf4, Tbx3, and LSD1. Cells transfected with miR-137 inhibitor and mock transfections were used as controls. RT-qPCR and Western Overexpression of miR-137 induces differentiation blot analyses showed that the mRNA and protein levels of of ES cells in vitro Klf4, Tbx3, and LSD1 were significantly reduced in cells transfected with mature miR-137 RNA duplexes compared to ES cells show the early stages of embryogenesis and can controls (Figs. 3A–C, and D). However, the pluripotency differentiate into all three primary embryonic germ layers; factor, Oct4, which lacks the miR-137 target site, remained the endoderm, mesoderm, and ectoderm; depending on unchanged in these cells (Fig. 3E). Cotransfection of the culture conditions (Evans and Kaufman, 1981). Therefore, mature miR-137 duplexes with the miR-137 antisense inhibitor we used EB cultures to determine the differentiation abolished the gene repression caused by miR-137 (Figs. 3A–C, potential of miR-137 in ES cells. We transfected the ES and D). These findings indicate that Klf4, Tbx3, and LSD1 are cells growing on gelatin-coated plates under self-renewing repressed by destabilization of their mRNAs by miR-137. These conditions with mature miR-137 RNA duplexes, miR-137 results are also consistent with the capacity of endogenous inhibitor, or a combination of both. After 24 h, cells were miR-137 in ES cells to repress ectopic 3′UTR reporters of Klf4, trypsinized and allowed to form EBs (D0) onto non-adherent Tbx3, and LSD1 (Figs. 2B, D, and F). We also observed the bacterial culture plates in ES medium without the leukemia downregulation of transcription factors Nanog and Sox2 in cells inhibitory factor (LIF) for 4 days (D4) as described previously transfected with miR-137 (Figs. 3F and G). However, the (Nair, 2006). Transfection of ES cells with miR-137 caused absenceofamiR-137targetsiteinNanogandSox2andthe changes in EB morphology, characterized by the appearance observation that Sox2 3′UTR luciferase reporter was not of a large number of flat cells with enlarged cytoplasm, even repressed by miR-137 (Fig. 3H) suggest that Sox2 is not a direct when the cells were maintained in complete ES medium target of miR-137, and the repression observed in these cells without any differentiation-inducing agents (Fig. 6A). The may be mediated by an indirect effect of miR-137. miR-137-induced morphological changes were completely prevented by co-transfection with miR-137 inhibitor (Fig. 6A). Overexpression of miR-137 modulates the To test whether differentiation of EBs by miR-137 was proliferation of R1 ES cells associated with the repression of Klf4, Tbx3, and LSD1, the expression of these genes was analyzed by RT-qPCR. Consistent The cell cycle-dependent expression of miR-137 in ES cells and with reported studies, the expression of ES cell markers Oct4, its upregulation during differentiation suggest a functional role Klf4, and Tbx3 was decreased in 4 day old EBs (D4) (Ivanova et for miR-137 in ES cell self-renewal as well as in differentiation. al., 2006; Martin-Ibanez et al., 2007; Sajini et al., 2012; Wei et To determine the role of miR-137 in cell proliferation, R1 ES al., 2005; Zhang et al., 2010). Moreover, in EBs transfected with cells growing under self-renewal conditions were transfected miR-137, there was significant downregulation of Klf4, Tbx3, with mature miR-137 duplexes, and cell proliferation was and LSD1, without affecting the expression of Oct4 (Figs. 6B–E). analyzed. In comparison to the control and miR-137 inhibitor These results suggest that the repression of Klf4, Tbx3, and transfected cells, miR-137 caused a significant drop in LSD1 by miR-137 correlates with the increased differentiation proliferation rate (Fig. 4A), which was abolished when cells of ES cells. Furthermore, the upregulation of neuronal (Sox1 were co-transfected with miR-137 inhibitor. We also investi- and Tuj1), endoderm (Foxa2 and Sox17), and mesoderm gated whether the maintenance of self-renewal under normal (Brachyury and Mixl1) markers in cultures of EBs transfected culturing conditions would be affected by increased miR-137 with mature miR-137 duplexes (Figs. 6F–K) suggests that by analyzing the self-renewal marker alkaline phosphatase. miR-137 accelerates the differentiation of ES cells into all The loss of alkaline phosphatase staining was evident from as three lineages. early as 2 days after transfection. In comparison to the control Previous studies have shown that miR-137 could inhibit and miR-137 inhibitor, transfection of mature miR-137 RNA proliferation and induce differentiation of tumor cells duplexes induced changes in colony morphology and caused a (Kozaki et al., 2008; Silber et al., 2008). Given that miR-137 is significant loss of alkaline phosphatase staining, indicative of essential for the maturation of neurons (Ripke et al., 2011; differentiation (Figs. 4B and C). These results indicate that Silber et al., 2008; Smrt et al., 2010; Sun et al., 2011a), we the significant drop in cell proliferation was caused by the next examined whether miR-137 induces differentiation of MicroRNA-epresses Klf4 and Tbx3 during differentiation of mouse embryonic stem cells 1307

Figure 4 Overexpression of miR-137 inhibits ES cell proliferation. (A) miR-137 significantly reduced cell proliferation. R1 ES cells transfected with mature miR-137 RNA duplexes and cultured in complete ES medium with LIF for 4 days. The growth rate was evaluated by CellTiter-Blue assay and the results are normalized to control ES cells (mean ± s.e.m., n = 8). (B) Alkaline phosphatase staining of control, mature miR-137 RNA duplexes, miR-137 inhibitor, and a combination of both. Transfected R1 ES cells after being cultured in complete ES medium with LIF for 5 days. (C) Relative numbers of alkaline phosphatase positive and negative colonies in panel B. Mean ± SD from triplicate measurements are plotted.

ES cells to a neuronal lineage. In EB differentiation assays, These data together suggest that miR-137 accelerates we stained the cells for the expression of the neuronal differentiation of ES cells by repressing its targets, including marker,Tuj1,andthemiR-137target,LSD1.Asshownin Klf4, Tbx3, and LSD1. Fig. 7, we observed a significant increase in the number of To further confirm the involvement of miR-137 in cells that express Tuj1 following transfection with miR-137 neurogenesis, we tested whether downregulation of miR-137 mimic. Moreover, the Tuj1-positive cells also showed could impair the differentiation of ES cells into neurons. To cellular morphology characteristics of differentiation (flat inactivate the endogenous miR-137 during differentiation, ES and elongated cells) and low levels of LSD1 (Figs. 7AandB). cells were transfected with miR-137 inhibitor or a control miR-137-induced morphological changes, the induction of inhibitor, allowed to form EBs, and then differentiated into Tuj1, and the repression of LSD1 were prevented when cells neurons by culturing for 12 days (Nair, 2006). Transfection of were cotransfected with miR-137 inhibitor (Figs. 7AandB). miR-137 inhibitor dramatically decreased the percentage of 1308 K. Jiang et al.

Figure 5 Overexpression of miR-137 represses the expression of self-renewal marker Nanog in ES cells. R1 ES cells were transfected with mature miR-137 RNA duplexes or miR-137 inhibitor and cultured in complete ES medium with LIF for 5 days. The cells were stained with antibodies to Nanog (red) and LSD1 (green) (A) or with Nanog (red) and cytochrome c (green) (B). Representative images were shown.

Tuj1-positive cells; Tuj1-positive staining was mainly observed transcription factors, Klf4 and Tbx3. It appears that miR-137 in rounded cells compared to the control and cells transfected has a dual function in ES cells depending on their expression with a control inhibitor (Figs. 8A and B). The differentiated status. While its expression in the mitotic phase indicates cells (NC) also expressed low levels of the Klf4, Tbx3, and that miR-137 is involved in cell division, increased levels of LSD1 proteins, compared to undifferentiated ES cells (ESC) the miRNA during differentiation appear to be involved in (Fig. 8C). To further determine the contribution of miR-137 in the maturation of ES cell-derived neurons. The data could the differentiation of ES cells to neurons, we assessed the indicate that the dynamic responses of miR-137 seem to be expression of ES cell and neuronal genes in differentiated cells involved in mediating the balance between self-renewal and transfected with miR-137 inhibitor by RT-qPCR. The ES cell differentiation by regulating the expression levels of its direct transcription factors were expressed at very low levels or targets, including Kfl4, Tbx3, and LSD1; and indirectly Nanog, were undetectable when ES cells were differentiated into Sox2, and possibly others. Our findings are consistent with the neuronal cells (Figs. 8D–H). In contrast, many of the neuronal reports that miRNAs promote mouse ES cell differentiation transporters and channels were highly expressed in differ- through posttranscriptional attenuation of key ES cell tran- entiated neurons (Figs. 8L–Q). The inhibition of miR-137 scription factors (Tay et al., 2008; Xu et al., 2009). significantly impaired the downregulation of Klf4, Tbx3, and LSD1 (Figs. 8G–I). A small but significant increase in the miR-137 is a known regulator of neuronal development levels of Nanog, Oct4, and Sox2 was also observed in and its expression has been found to increase during miR-137 inhibited cells (Figs. 8D–E). The increased levels of neuronal differentiation (Ripke et al., 2011; Silber et al., pluripotency factors in miR-137 inhibited cells were also 2008; Smrt et al., 2010; Sun et al., 2011a). This study correlated with decreased levels of neuronal genes such as demonstrates that miR-137 has a role in ES cell proliferation Calb1, Kcnn1, Scn3a, Slc1a3, and NeuroG2. No changes in the and differentiation. We have shown that miR-137 is levels of CoREST were observed in any of the treatment groups expressed in ES cells in a cell cycle-dependent manner and (Fig. 8K). This suggests that miR-137 is involved in the proper is highly upregulated during differentiation of ES cells into gene expression and maturation of neurons, at least in our in neuronal linage. These results suggest that miR-137 may be vitro differentiation system. involved in cell cycle progression or differentiation depending on the cellular context it is expressed in. This finding is consistent with previous reports, which show that Discussion miR-137 negatively regulates cell proliferation and accel- erates neural differentiation of embryonic neural stem cells In the present study, we have shown that miR-137, a (Silber et al., 2008; Sun et al., 2011b). miR-137 has been neuronally-enriched miRNA, is expressed in ES cells at the shown to mediate differentiation by targeting many genes mitotic phase of the cell cycle and its expression is highly including CDK6, a regulator of the cell cycle and dif- upregulated during differentiation into the neuronal linage. ferentiation (Silber et al., 2008); Ezh2, a histone We uncovered a direct link between miR-137 and ES cell methyltransferase (Szulwach et al., 2010); Jarid1b, a MicroRNA-epresses Klf4 and Tbx3 during differentiation of mouse embryonic stem cells 1309

Figure 6 Overexpression of miR-137 promotes differentiation of embryonic stem cells and represses Klf4, Tbx3, and LSD1. (A) Phase-contrast images showing the morphologies of EBs at day 4. Control (top left); miR-137 (top right); miR-137 inhibitor (bottom left); combination of both (bottom right). (B–E) RT-qPCR analysis of relative Klf4, Tbx3, LSD1, and Oct4 expression levels in EBs at days 0, 2, and 4. The R1 ES cells growing on gelatin coated plates were transfected with mature miR-137 duplexes, inhibitor, or a combination of both. After 24 h (Day 0), cells were dissociated and plated in a non-adherent plate to form EBs for 4 days (D4). The expression levels of each gene in the control at indicated days were arbitrarily set to one. The relative expression levels in EBs transfected with mature miR-137 RNA duplexes, miR-137 inhibitor, or a combination of both were normalized to the control mRNA levels of the corresponding day. The gene expression of Klf4 (B), Tbx3 (C), and LSD1 (D) was significantly repressed in miR-137 transfected cells, whereas the repression was completely prevented when cells were cotransfected with both miR-137 and miR-137 inhibitor. (F–K) RT-qPCR analysis of neuronal (Sox1 and Tuj1), endoderm (Foxa2 and Sox17), and mesoderm (Brachyury and Mixl1) markers in EBs transfected with mature miR-137 duplexes. All RT-qPCR experiments were performed three times in triplicate and the error bars denote SD. *, P b 0.001. histoneH3lysine4demethylase(Tarantino et al., 2010); and its expression has been shown to be downregulated during Mib1, a ubiquitin ligase (Smrt et al., 2010); and TLX, an ES cell differentiation (Adamo et al., 2011). In ES cells, we orphan nuclear (Sun et al., 2011b). In this study, found that miR-137 also targets the transcription factors, Klf4 we showed that miR-137 is also able to target the 3′UTRs of and Tbx3, in addition to LSD1. Transfection of mature miR-137 Klf4 and Tbx3 and regulate self-renewal and differentiation RNA duplexes into ES cells caused the repression of the Klf4, of ES cells. Tbx3, and LSD1 3′UTR assays; downregulation of endogenous Previously, LSD1 was identified as a miR-137 target in Klf4, Tbx3, and LSD1 mRNA and proteins; and triggered the neural stem cells of embryonic brains (Sun et al., premature differentiation in our EB assays. The cotransfection 2011b) and in colorectal cancer cells (Balaguer et al., of miR-137 and miR-137 inhibitor completely abolished the 2010). The interaction between miR-137 and LSD1 has been gene repression and differentiation mediated by the miR-137. shown to play a critical role in the molecular circuitry The inhibition of miR-137 impeded the dendritic morphogen- controlling neural stem cell proliferation and differentiation esis and phenotypic maturation of ES cell-derived neurons, during neural development (Sun et al., 2011b). Recent reports and the altered gene expression in these cells further suggests have demonstrated that high levels of LSD1 are needed to an important role for miR-137 in the differentiation of ES cells maintain the undifferentiated state of embryonic stem cells, in vitro. 1310 K. Jiang et al.

Figure 7 miR-137 promotes differentiation of ES cells toward a neuronal lineage. (A–B) Immunocytochemical staining demonstrates the differentiation of R1 ES cells transfected with the mature miR-137 RNA duplexes, miR-137 inhibitor, or a combination of both, 4 days after EB formation (5 days after transfection). Cells were stained for LSD1 (red) and Tuj1 (green). Lower (A) and higher (B) magnifications are shown. Scale bar, 250 and 40 μm, respectively.

Klf4 encodes a krűppel-like zincfinger , depending on the cellular context. The targeting of Klf4, Tbx3, which is required for the establishment of pluripotency in and LSD1 supports this notion, as Klf4, Tbx3, and LSD1 are somatic cells (Takahashi and Yamanaka, 2006). Tbx3 has been highly expressed in undifferentiated ES cells and in many forms reported to maintain ES cell self-renewal (Ivanova et al., 2006); of cancers (Fillmore et al., 2010; Hayami et al., 2011; Lim et and sustained expression of Tbx3 is sufficient to maintain ES al., 2010; Lomnytska et al., 2006; Lv et al., 2012; Renard et cells in an undifferentiated state in the absence of LIF (Niwa et al., 2007; Schulte et al., 2009; Vance et al., 2005; Wei et al., al., 2009). In mouse ES cells, the transcription factor, Klf4, is 2010). Recent evidence indicates that miR-137 represses LSD1 mainly activated by the Jak–Stat3 pathway and preferentially in human carcinoma cells and in mouse neural stem cells activates Sox2; whereas, Tbx3 is preferentially regulated by the as they begin to differentiate and is required for normal phosphatidylinositol-3-OH kinase–Akt and mitogen-activated differentiation (Balaguer et al., 2010; Sun et al., 2011b). Klf4 is protein kinase pathways and predominantly stimulates Nanog one of the four factors that, together, are sufficient to (Niwa et al., 2009; Zhang et al., 2010). However, activation of reprogram human fibroblasts into pluripotent cells (Takahashi Klf4 and Tbx3 is not required for the expression of Oct4 in ES and Yamanaka, 2006); and the overexpression of Tbx3 has cells (Niwa et al., 2009). We showed that miR-137 also been shown to increase the efficiency of the derivation of represses Nanog and Sox2 without affecting the expression of induced pluripotent stem cells (Han et al., 2010). It would be Oct4inthesecells.However,bothNanogandSox2lackmiR-137 interesting to determine, though, whether elevated miR-137 target sites; and miR-137 failed to repress Sox2 3′UTR activity. in mitotic cells is a factor of self-renewal or is involved in Our results suggest that miR-137 targets components of the differentiation, as demonstrated by the overexpression of self-renewal pathway in ES cells in order to promote differen- miR-137 in non-dividing cells. tiation by repressing Klf4 and Tbx3, which may indirectly downregulate Sox2 and Nanog. The downregulation of miR-137 in numerous cancers Conclusions (Balaguer et al., 2010; Chen et al., 2010; Silber et al., 2008), as well as our findings that miR-137 induces differentiation Here, we show that miR-137 is highly expressed in M phase ES of ES cells, raises the possibility that miR-137 functions as cells and shows increased expression in differentiated neuro- a pro-differentiation factor in a lineage-specific fashion nal cells. We also demonstrate that the increased expression MicroRNA-epresses Klf4 and Tbx3 during differentiation of mouse embryonic stem cells 1311

Figure 8 miR-137 participates in neuronal differentiation of mouse R1 ES cells. (A) Immunostaining for Tuj1 (green) and histone H3 (red) was performed at day 12 of differentiation, showing typical neurite formation in control (right panel) and cells transfected with a control inhibitor (left panel). The same analysis was performed in cells transfected with miR-137 inhibitor (middle panel) and shows a reduction in Tuj1 immunoreactivity, together with reductions in neurite outgrowth. Representative photomicrographs are shown. (B) Tuj1-positive cells in control; cells transfected with control inhibitor or miR-137 inhibitor were counted at day 12 of differentiation and expressed as percentage to control (mock-transfected) cells. The average of three culture replicates for each combination is shown, with error bar indicating the SD. *, P b 0.001. (C) Representative Western blots showing the levels of indicated proteins in ES cells and neuronally differentiated cells for 12 days. (D–Q) RT-qPCR analysis to examine the expression levels of ES cells and neuronal genes in cells transfected with control and miR-137 inhibitor at day 12 of differentiation. Transcript levels were normalized using β-actin, α-tubulin, and rps11 in these cells and compared with the levels in ES cells or neuronally differentiated ES cells for 12 days. All RT-qPCR experiments were performed two times in triplicate and the error bars represent SD. *, P b 0.001. 1312 K. Jiang et al. of miR-137 attenuates the levels of Klf4, Tbx3, and LSD1 Hayami, S., Kelly, J.D., Cho, H.S., Yoshimatsu, M., Unoki, M., during differentiation and its inhibition upregulates these Tsunoda, T., Field, H.I., Neal, D.E., Yamaue, H., Ponder, B.A., genes and disrupts neuronal maturation. Our results indicate Nakamura, Y., Hamamoto, R., 2011. Overexpression of LSD1 that miR-137 plays a role in the maintenance and differenti- contributes to human carcinogenesis through chromatin regula- tion in various cancers. Int. J. Cancer 128, 574–586. ation of ES cells. Ivanova, N., Dobrin, R., Lu, R., Kotenko, I., Levorse, J., DeCoste, C., Schafer, X., Lun, Y., Lemischka, I.R., 2006. Dissecting self- Author contributions renewal in stem cells with RNA interference. Nature 442, 533–538. Kim, J., Chu, J., Shen, X., Wang, J., Orkin, S.H., 2008. An extended V.D.N. planned, designed and supervised the project. K. J., transcriptional network for pluripotency of embryonic stem C. R., and V.D.N. carried out the experiments and analyzed cells. Cell 132, 1049–1061. the data. V.D.N. wrote the manuscript. Kozaki, K., Imoto, I., Mogi, S., Omura, K., Inazawa, J., 2008. Exploration of tumor-suppressive microRNAs silenced by DNA hypermethylation in oral cancer. Cancer Res. 68, 2094–2105. Conflicts of interest Lewis, B.P., Shih, I.H., Jones-Rhoades, M.W., Bartel, D.P., Burge, C.B., 2003. Prediction of mammalian microRNA targets. Cell 115, The authors declare no competing financial interests. 787–798. Li, M., Liu, G.H., Izpisua Belmonte, J.C., 2012. Navigating the epigenetic landscape of pluripotent stem cells. Nat. Rev. Mol. Acknowledgments Cell Biol. 13, 524–535. Lim, S., Janzer, A., Becker, A., Zimmer, A., Schule, R., Buettner, R., We thank Stuart Sealfon for helpful discussions, Kenneth Kosik Kirfel, J., 2010. Lysine-specific demethylase 1 (LSD1) is highly for Oct4 and Sox2 3′UTR constructs, and Sun Li for Tbx3 3′UTR expressed in ER-negative breast cancers and a biomarker predicting constructs. aggressive biology. Carcinogenesis 31, 512–520. Lomnytska,M.,Dubrovska,A.,Hellman,U.,Volodko,N., Souchelnytskyi, S., 2006. Increased expression of cSHMT, References Tbx3 and utrophin in plasma of ovarian and breast cancer patients.Int.J.Cancer118,412–421. Adamo, A., Sese, B., Boue, S., Castano, J., Paramonov, I., Barrero, Lv, T., Yuan, D., Miao, X., Lv, Y., Zhan, P., Shen, X., Song, Y., 2012. M.J., Izpisua Belmonte, J.C., 2011. LSD1 regulates the balance Over-expression of LSD1 promotes proliferation, migration and between self-renewal and differentiation in human embryonic invasion in non-small cell lung cancer. PLoS One 7, e35065. stem cells. Nat. Cell Biol. 13, 652–660. Marson, A., Levine, S.S., Cole, M.F., Frampton, G.M., Brambrink, Balaguer, F., Link, A., Lozano, J.J., Cuatrecasas, M., Nagasaka, T., T., Johnstone, S., Guenther, M.G., Johnston, W.K., Wernig, M., Boland, C.R., Goel, A., 2010. Epigenetic silencing of miR-137 is Newman, J., Calabrese, J.M., Dennis, L.M., Volkert, T.L., Gupta, an early event in colorectal carcinogenesis. Cancer Res. 70, S., Love, J., Hannett, N., Sharp, P.A., Bartel, D.P., Jaenisch, R., 6609–6618. Young, R.A., 2008. Connecting microRNA genes to the core Bartel, D.P., 2004. MicroRNAs: genomics, biogenesis, mechanism, transcriptional regulatory circuitry of embryonic stem cells. Cell and function. Cell 116, 281–297. 134, 521–533. Boyer, L.A., Lee, T.I., Cole, M.F., Johnstone, S.E., Levine, S.S., Martin-Ibanez,R.,Urban,N.,Sergent-Tanguy,S.,Pineda,J.R.,Garrido- Zucker, J.P., Guenther, M.G., Kumar, R.M., Murray, H.L., Clua, N., Alberch, J., Canals, J.M., 2007. Interplay of leukemia Jenner, R.G., Gifford, D.K., Melton, D.A., Jaenisch, R., Young, inhibitory factor and retinoic acid on neural differentiation of R.A., 2005. Core transcriptional regulatory circuitry in human mouse embryonic stem cells. J. Neurosci. Res. 85, 2686–2701. embryonic stem cells. Cell 122, 947–956. Melton, C., Judson, R.L., Blelloch, R., 2010. Opposing microRNA Calabrese, J.M., Seila, A.C., Yeo, G.W., Sharp, P.A., 2007. RNA families regulate self-renewal in mouse embryonic stem cells. sequence analysis defines Dicer's role in mouse embryonic stem Nature 463, 621–626. cells. Proc. Natl. Acad. Sci. U. S. A. 104, 18097–18102. Murchison, E.P., Partridge, J.F., Tam, O.H., Cheloufi, S., Hannon, Chambers, I., Tomlinson, S.R., 2009. The transcriptional foundation G.J., 2005. Characterization of Dicer-deficient murine embry- of pluripotency. Development 136, 2311–2322. onic stem cells. Proc. Natl. Acad. Sci. U. S. A. 102, 12135–12140. Chen, H., Shalom-Feuerstein, R., Riley, J., Zhang, S.D., Tucci, P., Nair, V.D., 2006. Activation of signaling initiates apoptotic death in Agostini, M., Aberdam, D., Knight, R.A., Genchi, G., Nicotera, a cellular model of Parkinson's disease. Apoptosis 11, 955–966. P., Melino, G., Vasa-Nicotera, M., 2010. miR-7 and miR-214 are Nair, V.D., Ge, Y., Balasubramaniyan, N., Kim, J., Okawa, Y., specifically expressed during neuroblastoma differentiation, Chikina, M., Troyanskaya, O., Sealfon, S.C., 2012. Involvement cortical development and embryonic stem cells differentiation, of histone demethylase LSD1 in short-time scale gene expression and control neurite outgrowth in vitro. Biochem. Biophys. Res. changes during cell cycle progression in embryonic stem cells. Commun. 394, 921–927. Mol. Cell. Biol. 32, 4861–4876. Dejosez, M., Zwaka, T.P., 2012. Pluripotency and nuclear Niwa, H., Ogawa, K., Shimosato, D., Adachi, K., 2009. A parallel reprogramming. Annu. Rev. Biochem. 81, 737–765. circuit of LIF signalling pathways maintains pluripotency of Evans, M.J., Kaufman, M.H., 1981. Establishment in culture of mouse ES cells. Nature 460, 118–122. pluripotential cells from mouse embryos. Nature 292, 154–156. Pauli, A., Rinn, J.L., Schier, A.F., 2011. Non-coding RNAs as Fillmore, C.M., Gupta, P.B., Rudnick, J.A., Caballero, S., Keller, regulators of embryogenesis. Nat. Rev. Genet. 12, 136–149. P.J., Lander, E.S., Kuperwasser, C., 2010. Estrogen expands Ramalho-Santos, M., Yoon, S., Matsuzaki, Y., Mulligan, R.C., Melton, breast cancer stem-like cells through paracrine FGF/Tbx3 D.A., 2002. “Stemness”: transcriptional profiling of embryonic signaling. Proc. Natl. Acad. Sci. U. S. A. 107, 21737–21742. and adult stem cells. Science 298, 597–600. Han, J., Yuan, P., Yang, H., Zhang, J., Soh, B.S., Li, P., Lim, S.L., Renard, C.A., Labalette, C., Armengol, C., Cougot, D., Wei, Y., Cao, S., Tay, J., Orlov, Y.L., Lufkin, T., Ng, H.H., Tam, W.L., Cairo, S., Pineau, P., Neuveut, C., de Reynies, A., Dejean, A., Lim, B., 2010. Tbx3 improves the germ-line competency of Perret, C., Buendia, M.A., 2007. Tbx3 is a downstream target of induced pluripotent stem cells. Nature 463, 1096–1100. the Wnt/beta-catenin pathway and a critical mediator of beta- MicroRNA-epresses Klf4 and Tbx3 during differentiation of mouse embryonic stem cells 1313

catenin survival functions in liver cancer. Cancer Res. 67, Takahashi, K., Yamanaka, S., 2006. Induction of pluripotent stem 901–910. cells from mouse embryonic and adult fibroblast cultures by Ripke, S., Sanders, A.R., Kendler, K.S., Levinson, D.F., Sklar, P., defined factors. Cell 126, 663–676. Holmans, P.A., Ln, D.Y., Duan, J., 2011. Genome-wide associ- Tarantino, C., Paolella, G., Cozzuto, L., Minopoli, G., Pastore, L., ation study identifies five new schizophrenia loci. Nat. Genet. Parisi, S., Russo, T., 2010. miRNA 34a, 100, and 137 modulate 43, 969–976. differentiation of mouse embryonic stem cells. FASEB J. 24, Sajini, A.A., Greder, L.V., Dutton, J.R., Slack, J.M., 2012. Loss of 3255–3263. Oct4 expression during the development of murine embryoid Tay, Y., Zhang, J., Thomson, A.M., Lim, B., Rigoutsos, I., 2008. bodies. Dev. Biol. 371, 170–179. MicroRNAs to Nanog, Oct4 and Sox2 coding regions modulate Schulte, J.H., Lim, S., Schramm, A., Friedrichs, N., Koster, J., embryonic stem cell differentiation. Nature 455, 1124–1128. Versteeg, R., Ora, I., Pajtler, K., Klein-Hitpass, L., Kuhfittig- Vance, K.W., Carreira, S., Brosch, G., Goding, C.R., 2005. Tbx2 is Kulle, S., Metzger, E., Schule, R., Eggert, A., Buettner, R., overexpressed and plays an important role in maintaining Kirfel, J., 2009. Lysine-specific demethylase 1 is strongly proliferation and suppression of senescence in melanomas. expressed in poorly differentiated neuroblastoma: implications Cancer Res. 65, 2260–2268. for therapy. Cancer Res. 69, 2065–2071. Wang, Y., Baskerville, S., Shenoy, A., Babiarz, J.E., Baehner, L., Silber, J., Lim, D.A., Petritsch, C., Persson, A.I., Maunakea, A.K., Blelloch, R., 2008. Embryonic stem cell-specific microRNAs Yu, M., Vandenberg, S.R., Ginzinger, D.G., James, C.D., regulate the G1-S transition and promote rapid proliferation. Costello, J.F., Bergers, G., Weiss, W.A., Alvarez-Buylla, A., Nat. Genet. 40, 1478–1483. Hodgson, J.G., 2008. miR-124 and miR-137 inhibit proliferation Wei, C.L., Miura, T., Robson, P., Lim, S.K., Xu, X.Q., Lee, M.Y., of glioblastoma multiforme cells and induce differentiation of Gupta, S., Stanton, L., Luo, Y., Schmitt, J., Thies, S., Wang, W., brain tumor stem cells. BMC Med. 6, 14. Khrebtukova, I., Zhou, D., Liu, E.T., Ruan, Y.J., Rao, M., Lim, B., Silva, J., Smith, A., 2008. Capturing pluripotency. Cell 132, 532–536. 2005. Transcriptome profiling of human and murine ESCs Smrt, R.D., Szulwach, K.E., Pfeiffer, R.L., Li, X., Guo, W., Pathania, identifies divergent paths required to maintain the stem cell M., Teng, Z.Q., Luo, Y., Peng, J., Bordey, A., Jin, P., Zhao, X., state. Stem Cells 23, 166–185. 2010. MicroRNA miR-137 regulates neuronal maturation by Wei, D., Wang, L., Kanai, M., Jia, Z., Le, X., Li, Q., Wang, H., Xie, targeting ubiquitin ligase mind bomb-1. Stem Cells 28, 1060–1070. K., 2010. KLF4alpha up-regulation promotes cell cycle progres- Stefani, G., Slack, F.J., 2008. Small non-coding RNAs in animal sion and reduces survival time of patients with pancreatic development. Nat. Rev. Mol. Cell Biol. 9, 219–230. cancer. Gastroenterology 139, 2135–2145. Sun, G., Ye, P., Murai, K., Lang, M.F., Li, S., Zhang, H., Li, W., Fu, Xu, N., Papagiannakopoulos, T., Pan, G., Thomson, J.A., Kosik, C., Yin, J., Wang, A., Ma, X., Shi, Y., 2011a. miR-137 forms a K.S., 2009. MicroRNA-145 regulates OCT4, SOX2, and KLF4 and regulatory loop with TLX and LSD1 in neural represses pluripotency in human embryonic stem cells. Cell 137, stem cells. Nat. Commun. 2, 529. 647–658. Sun, S.G., Zheng, B., Han, M., Fang, X.M., Li, H.X., Miao, S.B., Su, Zhang, P., Andrianakos, R., Yang, Y., Liu, C., Lu, W., 2010. Kruppel- M., Han, Y., Shi, H.J., Wen, J.K., 2011b. miR-146a and Kruppel- like factor 4 (Klf4) prevents embryonic stem (ES) cell differen- like factor 4 form a feedback loop to participate in vascular tiation by regulating Nanog gene expression. J. Biol. Chem. 285, smooth muscle cell proliferation. EMBO Rep. 12, 56–62. 9180–9189. Szulwach, K.E., Li, X., Smrt, R.D., Li, Y., Luo, Y., Lin, L., Santistevan, Zhang, Z., O'Rourke, J.R., McManus, M.T., Lewandoski, M., Harfe, N.J.,Li,W.,Zhao,X.,Jin,P.,2010.Cross talk between microRNA B., D. Sun, X., 2011. The microRNA-processing enzyme Dicer is and epigenetic regulation in adult neurogenesis. J. Cell Biol. 189, dispensable for somite segmentation but essential for limb bud 127–141. positioning. Dev. Biol. 351, 254–265.