Neural stem cells derived from epiblast stem cells display distinctive properties

Stem Cell Research (2014) 12, 506–516

Available online at www.sciencedirect.com

ScienceDirect www.elsevier.com/locate/scr

Neural stem cells derived from epiblast stem cells display distinctive properties☆ Hyo Jin Jang a,1 , Jong Soo Kim a,1 , Hyun Woo Choi a,1 , Iksoo Jeon b , Sol Choi a , Min Jung Kim a , Jihwan Song b , Jeong Tae Do a,⁎ a

Department of Animal Biotechnology, College of Animal Bioscience and Technology, Konkuk University, Seoul 143–701, Republic of Korea b Department of Biomedical Science, CHA University, 606–16 Yeoksam 1-dong, Kangnam-gu, Seoul 135–081, Republic of Korea

Received 13 June 2013; received in revised form 24 December 2013; accepted 26 December 2013 Available online 4 January 2014

Abstract Pluripotent stem cells can be derived from preimplantation and postimplantation mouse embryos. Embryonic stem cells (ESCs) derived from blastocysts are in a “naive” pluripotent state and meet all of the criteria for pluripotency, including the ability to generate live pups through tetraploid complementation. Epiblast stem cells (EpiSCs) derived from postimplantation epiblasts are in a “primed” pluripotent state. ESCs and EpiSCs show different phenotypes and gene expression patterns, and EpiSCs are thought to be less pluripotent than ESCs. In this study, we addressed whether EpiSCs can be differentiated into specialized cell types in vitro. To do this, we first derived EpiSCs from E5.5–6.5 mouse embryos containing the Oct4-GFP transgene. We found that EpiSCs expressed pluripotency markers and differentiated into all three germ layers in intro and in vivo. Interestingly, EpiSCs also efficiently differentiated into a homogenous population of neural stem cells (NSCs) in vitro. The EpiSC-derived NSCs (EpiSC-NSCs) expressed NSC markers (Nestin, Sox2, and Musashi), self-renewed for more than 20 passages, and differentiated into neuronal and glial neural cell subtypes in vitro. We then transplanted the EpiSC-NSCs into the neonatal mouse brains, and found that they were able to survive and differentiate into robust neurons and glial cells in the mouse brains, demonstrating that primed pluripotent EpiSCs efficiently form functional NSCs. We compared the global gene expression patterns of NSCs differentiated from EpiSC-NSCs, ESCs, and brain tissue and found that the expression patterns of most genes, including pluripotency and NSC specificity, were similarly clustered, but that the developmental process-related genes were distantly clustered. Moreover, the global gene expression pattern of brain-derived NSCs was more similar to that of ESC-derived NSCs than that of EpiSC-derived NSCs. Taken together, these results indicate that although NSCs, regardless of their origins, display very similar in vitro and in vivo differentiation properties, their global gene expression profiles may differ, depending on the pluripotency state, i.e., naive or primed. © 2014 The Authors. Published by Elsevier B.V. All rights reserved.

☆ This is an open-access article distributed under the terms of the Creative Commons Attribution–Noncommercial–No Derivative Works License, which permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are credited. ⁎ Corresponding author. Fax: + 82 2 455 1044. E-mail address: [email protected] (J.T. Do). 1 These authors contributed equally to this work.

Introduction Pluripotent stem cells can be classified into two distinct pluripotent states, i.e., “naive” and “primed.” Cells of the inner cell mass (ICM) of the blastocyst and epiblast cells are the pluripotent cells in early embryonic development.

1873-5061/$ - see front matter © 2014 The Authors. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.scr.2013.12.012

507 Embryonic stem cells (ESCs) derived from ICM of blastocyst are naive pluripotent (Nichols and Smith, 2009) and epiblast stem cells (EpiSCs) derived from postimplantation epiblast cells are primed pluripotent. Naive pluripotent mouse ESCs are maintained in a pluripotent state by signaling through the STAT3 and bone morphogenetic protein 4 (BMP4) pathways. These cells contribute to all three germ layers (including germ cells) in vitro and in vivo and efficiently form chimeric mice (Meissner et al., 2008; Nichols and Smith, 2009; Ying et al., 2008). By contrast, the primed pluripotent EpiSCs are maintained in a pluripotent and self-renewing state by signaling through basic fibroblast growth factor (bFGF) and Activin/Nodal (Brons et al., 2007; Vallier et al., 2009); however, EpiSCs barely contribute to chimeras when injected into blastocysts, the efficiency of chimera formation being as low as 0.5% (2/385) (Brons et al., 2007). EpiSCs and ESCs also differ in other aspects, including colony morphology (Stadtfeld and Hochedlinger, 2010), specific marker expression patterns (Guo et al., 2009; Han et al., 2011; Jouneau et al., 2012), and the number of active X chromosomes in female cells (Gardner et al., 1985). Human ESCs (hESCs) derived from preimplantation blastocysts are more similar to mouse EpiSCs than to mouse ESCs in terms of their flat morphology, signaling requirements for maintenance of pluripotency, and X chromosome inactivation status (Aoki et al., 2009; Hanna et al., 2010a; Silva et al., 2008). However, the state of pluripotency is not stringent, because EpiSCs can be produced from blastocysts in culture (Najm et al., 2011a), and EpiSCs and hESCs can be converted into naive pluripotent state as mouse ESCs by genetic manipulation or treatment with small molecules (Gillich et al., 2012; Hanna et al., 2010a, 2010b). EpiSCs could be differentiated into all three germ layers in vitro and in vivo; however, it is not yet known whether EpiSCs can differentiate into specialized cell types in vitro. Recently, Najm et al. showed that EpiSCs could differentiate into oligodendrocyte progenitor cells that subsequently upregulated oligodendrocyte-specific genes and differentiated into oligodendrocytes in vitro (Najm et al., 2011b). However, it is not yet known whether EpiSCs are able to differentiate into NSCs in vitro. In this study, we confirm that EpiSCs can be differentiated into NSCs in vitro. We show that EpiSC-derived NSCs expressed the NSC markers Nestin, Sox2, and Musashi, and self-renewed for more than 20 passages in vitro. EpiSC-derived NSCs were indistinguishable from brain-derived NSCs and ESC-derived NSCs in their morphological characteristics, differentiation potential, and NSC-related gene expression. We also examined the global gene expression profiles of NSCs derived from EpiSC, ESC, and brain.

Materials and methods Culture of EpiSC Epiblasts were obtained from OG2 heterozygous embryos at embryonic day 6.5 (E6.5). OG2 mice express GFP under the control of the Oct4 promoter and a distal enhancer. All mouse strains were bred and housed at the mouse facility of Konkuk University or were bought from The Jackson Laboratory (Sacramento, CA, USA). Animal handling was

done in accordance with the animal protection guidelines of the Konkuk University and the Korean animal protection laws. The decidua were removed from the pregnant mouse with a needle, and the extraembryonic ectoderm was separated from the epiblast with hand-pulled glass pipettes. The epiblast was washed with PBS and cultured on a layer of inactivated mouse embryonic fibroblasts (MEFs) in EpiSC medium (Brons et al., 2007), consisting of DMEM/F12 supplemented with 20% KnockOut Serum Replacement (both from Gibco), 2 mM glutamine, nonessential amino acids, 20 ng/ml activin A (PeproTech), and 5 ng/ml basic fibroblast growth factor (bFGF; Invitrogen). EpiSCs were passaged every 3 to 4 days by dissociation with 1 mg/ml collagenase type IV (Invitrogen) or by pipetting with a glass pipette.

Derivation of NSCs from EpiSCs For NSC induction, EpiSCs were cultured in neural induction medium (DMEM/F12 [Invitrogen] supplemented with 20% KSR [Invitrogen], 20 ng/ml activin A, and 12 ng/ml bFGF). After 1 to 2 days, differentiated cells (identified by their change in morphology) were transferred into the neurosphere medium consisting of DMEM/F12 supplemented with N2 and B27 supplements (both Invitrogen), 2 mM glutamine, 1× penicillin/streptomycin/glutamine (Invitrogen), 20 ng/ml EGF (Invitrogen), and 20 ng/ml bFGF, and cultured for 2 days in suspension culture. Cells were then resuspended in NSC expansion medium (DMEM/F12 supplemented with N2 supplement, 10 ng/ml EGF, and bFGF, 50 μg/ml bovine serum albumin [Invitrogen], and 1× penicillin/streptomycin/ glutamine) and cultured in a 0.1% gelatin-coated dish for 3 days.

Differentiation of NSCs into neurons and glial cells Brain-, ESC-, or EpiSC-derived NSCs were differentiated into neurons and glial cells by culturing for 3 weeks in neuronal differentiation medium (DMEM/F12 medium, Neurobasal medium [Invitrogen], 1% fetal bovine serum, 1 × N2 supplement, 1 × penicillin/streptomycin/glutamine, and 50 × B27 supplement).

Alkaline phosphatase assay For alkaline phosphatase (AP) staining, cells were fixed with 4% paraformaldehyde for 1 min at room temperature, washed with 0.05% Tween-20/PBS, and treated with AP solution (Takara) for 15 min at RT.

Immunocytochemistry Cells were fixed with 4% paraformaldehyde for 20 min at room temperature, washed with PBS, resuspended in PBS containing 10% normal goat serum and 0.03% Triton X-100, and treated with primary antibodies for 45 min at room temperature. The primary antibodies were anti-Oct4 (1:500 dilution; Chemicon, Temecula, CA, USA), anti-Nanog (1:250; Chemicon), antiNestin (1:200; Chemicon), anti-Sox2 (1:1000; Chemicon), anti-Musashi (1:1000; Chemicon), anti-β-III tubulin (Tuj1, 1:1000; Chemicon), and anti-glial fibrillary acidic protein

508 (GFAP, 1:250; Chemicon). Cells were then incubated with fluorescence-labeled secondary antibodies (Alexa Fluor 488 or 568; Molecular Probes, Eugene, OR, USA), according to the manufacturer's specifications.

Teratoma formation To confirm their ability to undergo differentiation in vivo, EpiSCs (~ 1 × 106 cells) were injected into the testes of SCID mice. Tumors were harvested surgically from the euthanized mice at 4 weeks after injection. Teratomas were fixed with 4% paraformaldehyde for 36 h at RT and then sectioned using standard techniques. Sections were stained with hematoxylin and eosin to identify the tissue types.

RNA isolation and quantitative reverse-transcriptase polymerase chain reaction (RT-PCR) Total RNA was isolated using the RNeasy Mini Kit (Qiagen) and treated with DNase to remove contaminating genomic DNA. Samples of 1 μg total RNA were reverse transcribed with a SuperScript III Reverse-Transcriptase Kit and oligo (dT) primers (both Invitrogen) according to the manufacturer's instructions. The primer sequences and annealing temperatures were as follows: Oct4 sense 5′-TCTCCCATGCATTCA AACTG and antisense 5′-GCTCCTGATCAACAGCATCA; Nanog sense 5′-AAGTACCTCAGCCTCCAGCA and antisense 5′-GTGCTG AGCCCTTCTGAATC; FGF5 sense 5′-GCGACGTTTTCTTCGTCTTC and antisense 5′-ACAATCCCCTGAGACACAGC; Sox17 sense 5′-GGAGGGTCACCACTGCTTTA and antisense 5′-AGATGTC TGGAGGTGCTGCT; Dkk1 sense 5′-GAGGGGAAATTGAGGAA AGC and antisense 5′-GGTGCACACCTGACCTTCTT; and Rex1 sense 5′-CTGGTGGCAAGTATGTGGTG and antisense 5′-CAG GAGTGTATGGCTCAGCA. Quantitative PCR reactions were performed in duplicate using Power SYBR Green Master Mix (Takara) and analyzed with a Roche LightCycler 5480 (Roche). Duplicate amplifications were carried out for each target gene with three wells serving as negative controls. Quantification was estimated using the comparative Ct method normalized with β-actin and presented as a percentage of biological controls.

H.J. Jang et al. and sequenced. BiQ Analyzer software (Max Planck Institute, Saarbrücken, Germany) was used to visualize and quantify the bisulfite sequence data.

Transplantation of brain-, ESC-, and EpiSC-derived NSCs NSC transplantation was performed as described previously (Deng et al., 2003, 2006). In brief, NSCs were stained with bisbenzimide and transplanted into the lateral ventricle of wild-type neonatal CF1 mice within 24 h of birth. For transplantation, the pups were anesthetized by hypothermia and placed in a clay mold. The head was transilluminated under a dissection microscope, and a Hamilton syringe with a beveled tip was lowered through the scalp and skull immediately anterior to the bregma. Approximately 2.5 × 105 NSCs in 1 μl of DMEM/F12 were then slowly injected into the left lateral ventricle under low pressure. Immediately after injection, pups were warmed in a 37 °C incubator and returned to the mother approximately 30 min later. At 10 days posttransplantation, mice were euthanized with an overdose of Avertin and then transcardially perfused with 4% paraformaldehyde in PBS. The brain was excised, postfixed overnight in 4% paraformaldehyde in PBS, and sectioned through the coronal plane into 40-μm slices with a cryostat (Microm, Walldorf, Germany).

Immunolabeling of brain sections Forebrains were cut with a cryotome into 40-μm coronal sections exhaustively and processed free-floating. After blocking in PBS with 10% goat serum, sections were incubated at 4 °C in primary antibodies against the neuron-specific marker Tuj1(type III beta-tubulin; 1:500, BD Bioscience) or the astrocyte-specific marker GFAP (glial fibrillary acidic protein; 1:500, BD Bioscience). The tissues were then washed in PBS and incubated in the appropriate fluorescence-conjugated secondary antibodies (goat anti-rabbit IgG-conjugated Alexa488, 1:200, Molecular Probes) for 1 h at room temperature. After a final wash in PBS, the stained brain slices were mounted onto glass slides, examined and photographed using a confocal laser-scanning microscope imaging system (LSM510, Carl Zeiss, Inc.).

Methylation analysis Microarray-based analysis Genomic DNA was isolated from ESCs and EpiSCs and bisulfite reactions were performed using an EpiTect Bisulfite Kit (Qiagen) according to the manufacturer's instructions. PCR amplification of Stella and Dppa5 promoter regions was performed using the following primer sequences and annealing temperatures: Stella first primer pair, sense 5′-TTTTTT TATTTTGTGATTAGGGTTG and antisense 5′-CTTCACCTAAAC TACACCTTTAAAC (161 bp, 45 °C); Stella second primer pair, sense 5′-TTTGTTTTAGTTTTTTTGGAATTGG and antisense 5′-CT TCACCTAAACTACACCTTTAAAC (116 bp, 55 °C); Dppa5 first primer pair, sense 5′-GGTTTGTTTTAGTTTTTTTAGGGGTATA and antisense 5′-CCACAACTCCAAATTCAAAAAAT (140 bp, 45 °C); Dppa5 second primer pair, sense 5′-TTTAGTTTTT TTAGGGGTATAGTTTG and antisense 5′-CTTCACCTAAAC TACACCTTTAAAC (132 bp, 55 °C). The PCR products were subcloned into pGEM-T Easy vector (Promega, Madison, WI)

Total RNA was isolated using an RNeasy Mini Kit (Qiagen) and digested with DNase I (RNase-free DNase, Qiagen) according to the manufacturer's instructions. Total RNA was amplified, biotinylated, and purified using the Ambion Illumina RNA amplification kit (Ambion), according to the manufacturer’s instructions. Biotinylated cRNA samples (750 ng) were hybridized to each MouseRef-8 v2 Expression BeadChip (Illumina) and signals were detected using Amersham fluorolink streptavidinCy3 (GE Healthcare Life Sciences) by following the Illumina bead array protocol. Arrays were scanned with an Illumina BeadArray Reader confocal scanning system according to the manufacturer's instructions. Raw data were extracted using the software provided by the manufacturer (Illumina GenomeStudio v2011. 1, Gene Expression Module v1.9. 0). Array data were filtered using a

509 detection p-value of b 0.05 in at least 50% of the samples. The selected probe signal value was logarithmically transformed and normalized using the quantile method. Comparative analyses were carried out using the LPE test and fold change. The false discovery rate was controlled by adjusting the p-values using the Benjamini–Hochberg algorithm. Hierarchical clustering was performed using average linkage and Pearson distance as a measure of similarity. Gene Ontology analysis (Functional Annotation Clustering) was performed using the DAVID software (Huang da et al., 2009), with the classification stringency set to high.

Results EpiSCs derived from OG2 E6.5 embryos are in a primed pluripotent state EpiSCs were established from E6.5 postimplantation embryos of OG2 heterozygous mice. OG2 mice express GFP under the

control of the Oct4 promoter and distal enhancer. EpiSCs grew on a MEF feeder layer as flat colonies reminiscent of hESCs (Fig. 1A). Oct4 in EpiSCs is regulated by the promoter and proximal enhancer, which is missing from the Oct4-GFP regulatory sequence of OG2 EpiSCs (Han et al., 2010). Thus, epiblast cells expressed Oct4-GFP in the initial stages of culture, but expression was extinguished during subculture (Fig. 1A). To determine whether the OG2 EpiSCs exist in a primed pluripotent state, as reported for other EpiSC lines, we analyzed the expression of alkaline phosphatase (AP) activity and pluripotency markers, the DNA methylation status of germ cell marker genes, global gene expression levels, and in vitro and in vivo differentiation potential. EpiSCs were very weakly positive for AP compared with the strong staining exhibited by mouse ESCs (Fig. 1B). EpiSCs were positive for Oct4 expression by immunofluorescence staining, whereas Nanog was barely detected (Fig. 2A). Next, we performed bisulfite genomic sequencing to assess the DNA methylation status of the germ cell marker Stella and the early epiblast (naive pluripotency) marker Dppa5. We found that the promoter regions of Stella

Figure 1 Generation of epiblast stem cells (EpiSCs) from E6.5 postimplantation Oct4-GFP transgenic embryos. (A) Epiblast tissue was dissected from E6.5 mouse embryos and cells were cultured on the inactivated mouse embryonic fibroblasts. EpiSCs grew on the feeder layer as flat colonies. (B) Alkaline phosphatase (AP) staining was very weak in EpiSCs and strong in embryonic stem cells.

510 and Dppa5 were unmethylated in naive pluripotent mouse ESCs but hypermethylated in EpiSCs (Fig. 2B). Real-time RT-PCR analysis also revealed that expression of Oct4 mRNA was almost the same in EpiSCs and ESCs, whereas Nanog mRNA levels were lower in EpiSCs than in ESCs. EpiSCs expressed high levels of the epiblast markers FGF5, Sox17 and Dkk1 and much lower levels of the ESC-specific marker, Rex1 (Fig. 2C). Upon

H.J. Jang et al. injection into SCID mice testes, EpiSCs formed teratomas containing various tissues, including gut (endoderm), muscle (mesoderm), and neural (ectoderm) tissues, indicating that the EpiSCs could differentiate into all three germ layers in vivo (Fig. 2D). Taken together, these data show that EpiSCs derived from E6.5 OG2 epiblasts meet all the criteria for primed pluripotency.

Figure 2 Characterization of epiblast stem cells (EpiSCs). (A) Immunofluorescence analyses of pluripotency markers showed EpiSCs were strongly positive for Oct4 but very weakly positive for Nanog. (B) DNA methylation status of the promoter regions of Stella and Dppa5 in EpiSCs and naive pluripotent embryonic stem cells (ESCs) was analyzed by bisulfite genomic sequencing analysis. (C) Real-time RT-PCR analysis of EpiSCs and ESCs for pluripotency-specific (Oct4, Nanog, and Rex1) and EpiSC-specific (FGF5, Sox17, Dkk1) markers. Real-time RT-PCR data are normalized to β-actin expression. (D) Teratoma analysis from mouse EpiSCs. Derivatives of all three germ layers were present; endoderm (gut epithelium), mesoderm (muscle fibers), and ectoderm (neural rosettes). Scale bars 100 μm.

511

Derivation of NSCs from EpiSCs To determine whether EpiSCs, like ESCs, could differentiate into specific somatic stem cell lineages in vitro, we induced them into NSCs. NSCs can be maintained in culture as pure self-renewing stem cells for an extended period (Chojnacki et al., 2009; Temple, 2001). EpiSCs were differentiated into NSCs using a modification of previous methods (Conti et al., 2005). EpiSCs were cultured in a neural induction medium with high concentration of bFGF (15 ng/ml) for 1 to 2 days and then transferred into neurosphere medium on non-adherent culture dishes for 2 days to induce neurosphere formation (Fig. 3A). During neural induction, EpiSCs changed their morphology and formed neurosphere-like aggregates in suspension culture (Fig. 3B). The neurosphere-like aggregates were transferred into NSC culture medium on gelatin-coated dishes and cultured for ~1 week, at which point they form NSCs (Figs. 3A and B). EpiSC-derived NSCs (EpiSC-NSCs) were morphologically indistinguishable from brain-derived NSCs (Fig. 3B), could be propagated more than passage 20 and showed positive staining for the NSC markers, Sox2, Nestin, and Musashi (Fig. 3C). These results therefore suggested that primed pluripotent EpiSCs could directly differentiate into NSCs in vitro, and the EpiSCderived NSCs showed similarities to brain-derived NSCs.

Neuronal and glial differentiation potency of EpiSC-NSCs in vitro and in vivo NSCs are multipotent stem cells that possess the ability to differentiate into neurons and glial cells (astrocytes

and oligodendrocytes) (Gage, 2000). To determine whether EpiSC-NSCs can differentiate into subtypes of cells in vitro, we incubated NSCs derived from EpiSCs, brain tissue (NSCs), or ESCs (ESC-NSCs) in neural differentiation medium (without growth factors) for ~ 3 weeks. Oligodendrocytes were not observed with this differentiation protocol. However, upon staining with antibodies to GFAP (astrocyte marker) or Tuj1, Map2 (neuronal marker) (Fig. S1A), we found that cells from all three differentiated cultures contained cells positive for GFAP (~50%) and Tuj1 (~10%) (Figs. 4A and B). About 40% of derivatives from NSCs might be undifferentiated cells, which were still Sox2 positive (Fig. S3). These results indicated that EpiSC-NSCs have the ability to differentiate into neuronal and glial cells in vitro with the same efficiency as the control NSC and ESC-NSC lines. Next, we tested whether the EpiSCNSCs were able to survive and differentiate into functional neurons and glial cells after transplantation into postnatal mouse brains. One to 4 weeks after transplantation, we examined the brains, especially where bisbenzimide-positive signals were detected. Unlike human ESCs or NSCs, in which transplanted cells can be easily detected following transplantation by immunohistochemistry using antibodies against human-specific nuclei or mitochondria, mouse-derived cells are known to be very difficult to detect using species-specific antibodies. To overcome this problem, mouse-derived cells have been pre-labeled with various kinds of substances, although they have intrinsic problems of diffusion or dilution during cell division. In this study, we utilized bisbenzimide, a fluorescent dye for staining DNA in the nuclei, which we have routinely used for the detection of transplanted cells. Although bisbenzimide-positive staining cannot guarantee the

Figure 3 Generation of neural stem cells (NSCs) from epiblast stem cells (EpiSCs). (A) Timeline for the differentiation of EpiSCs into NSCs. (B) Morphology of neurospheres and NSCs differentiated from EpiSCs. EpiSC-derived NSCs self-renew for more than 20 passages in the presence of epidermal growth factor and basic fibroblast growth factor. (C) Expression of NSC markers (Sox2, Nestin, and Musashi) in EpiSC-derived NSCs.

512

H.J. Jang et al.

Figure 4 Differentiation potential of epiblast stem cells (EpiSCs) into neurons and glial cells in vitro and in vivo. (A) Brain-, embryonic stem cell (ESC)-, and EpiSC-derived neural stem cells (NSCs) differentiated into neurons and glial cells in vitro, as shown by immunofluorescence staining of the neuronal marker, Tuj1, and the glial (astrocyte) marker, glial fibrillary acidic protein (GFAP). (B) Quantification of the lineage-specific differentiation efficiency by brain-, ESC-, and EpiSC-derived NSCs. Error bars indicate the standard error of the mean. (C) In vivo differentiation of brain-, ESC-, and EpiSC-derived NSCs following implantation into neonatal mouse brains. Immunohistochemical staining indicates that NSCs derived from brain (top), ESCs (middle), and EpiSC (bottom) were positive for the neuronal marker (Tuj1) and the glial (astrocyte) marker (GFAP). Derivatives of implanted NSCs were distinguished by bisbenzimide staining. Scale bars 20 μm.

monitoring of transplanted cells with 100% accuracy (Iwashita et al., 2000), they can serve as a useful guideline to determine the location of transplanted cells. Having this in mind, we observed that transplanted NSCs could migrate into multiple

sites of brain and differentiate into neurons and glial cells (Fig. S4 and S5). The migrated cells were detected as large colonies, indicating that bisbenzimide-positive cells were not just an artifact caused by diffusion of bisbenzimide. We also

Figure 5 Global gene expression profiles and Gene Ontology analysis of pluripotent and differentiated cells. (A) Heat map of global gene expression in embryonic stem cells (ESCs), epiblast stem cells (EpiSCs), mouse embryonic fibroblasts, and EpiSC-, brain-, and ESC-derived neural stem cells (NSCs). (B) Hierarchical clustering analysis of the global gene expression profiles using the average linkage and the Pearson distance. (C) Heat map of genes upregulated in EpiSC-NSCs compared with NSCs. (D) Heat map of genes downregulated in EpiSC-NSCs compared with NSCs. (E) Gene Ontology biological process terms of the genes upregulated in EpiSCNSCs. (F) Gene Ontology biological process terms of the genes downregulated in EpiSC-NSCs. Functional annotation clustering was performed using the DAVID algorithm.

513

514 found that transplanted EpiSC-NSCs had survived and differentiated into neurons and astrocytes (Fig. 4C) as well as the transplanted brain-derived and ESC-derived NSCs. Taken together, these results demonstrate that EpiSC-derived NSCs can differentiate into neurons and glial cells in vitro and in vivo with potency comparable to that of brain-derived and ESC-derived NSCs.

Analysis of global gene expression in EpiSC-NSCs To identify differences in the molecular signatures of NSCs, ESC-NSCs, and EpiSC-NSCs, we analyzed their gene expression profiles using microarrays (Illumina MouseRef-8 v2 Expression BeadChip). The microarray data showed that the pattern of global gene expression of EpiSC-NSCs was clearly distinct from that of pluripotent ESCs, EpiSCs, or MEFs, another type of differentiated somatic cell (Figs. 5A and B). Although the expression patterns of EpiSC-NSCs and NSCs were very similar, there were differences in several gene clusters. To identify the gene expression pattern that was unique to EpiSC-NSCs, and not found in NSCs, we looked for genes enriched in Gene Ontology biological processes by screening for transcripts with a 3-fold change in expression. We found that 114 genes were upregulated and 135 genes were downregulated in EpiSC-NSCs (Figs. 5C and D). The Gene Ontology categories most significantly enriched in EpiSC-NSCs were “multicellular organismal development,” “cell development,” “generation of neurons,” “neuron differentiation,” “transmission of nerve impulse,” and “blood vessel morphogenesis,” which had ≥ 1.3 enrichment scores with the classification stringency set to high (Fig. 5E). On the other hand, the most significantly downregulated transcripts in EpiSC-NSCs were “anatomical structure development,” “regulation of response to stimulus,” “regulation of cell death,” “positive regulation of cell death,” “chemotaxis,” “positive regulation of cytokine production,” “response to molecule of bacterial origin,” “immune response-activating signal transduction,” “toll-like receptor signaling pathway,” and “positive regulation of adaptive immune response,” which had ≥ 1.3 enrichment scores with the classification stringency set to high (Fig. 5F). Of the differentially expressed genes, only 9 were upregulated and 33 were downregulated in ESCNSCs compared with NSCs, and these could not be categorized by Gene Ontology (data not shown). The Gene Ontology cluster “developmental process” was common to the upregulated and downregulated gene sets in EpiSC-NSCs compared with NSCs, indicating that the differentiation process of EpiSCs into NSCs may differ from that of ESCs into NSCs (Supplementary Table A and B). Nevertheless, the expression profiles of most genes that are related to pluripotency, EpiSC specificity, and NSC specificity were similarly clustered between the cell types (Fig. S2). Given that EpiSC-NSCs and NSCs showed the same ability to self-renew and the potential for differentiation into neural subtypes in vivo and in vitro, the differences in gene expression might not affect the function of NSCs. Taken together, our results suggest that pluripotent stem cell-derived NSCs express the same NSC markers and have the same potential to differentiate into neural subtypes as brain-derived NSCs. Although NSCs show differences in their global gene expression patterns depending on the pluripotent stem cell of origin (naive or primed), these may not affect NSC differentiation potential.

H.J. Jang et al.

Discussion Here, we showed that primed pluripotent stem cells, EpiSCs, were able to differentiate into highly expandable tissue-specific stem cells (NSCs) in vitro. These EpiSC-NSCs were morphologically similar to brain-derived NSCs, expressed NSC markers (Nestin, Sox2, and Musashi), and could differentiate into neurons and glial cells both in vitro and in vivo. In this respect, EpiSC-NSCs have multipotent differentiation potential comparable to that of brain-derived NSCs. However, the global gene expression patterns of the differentiated derivatives of the primed pluripotent EpiSCs and naive pluripotent ESCs were not the same, with the gene expression profile of ESC-derived NSCs being closer to that of brain-derived NSCs. These results suggest that, for in vitro differentiated cells, the entire transcriptome of naive pluripotent stem cell-derived NSCs (and perhaps all cell types) might be closer to that of tissue-derived cells than cells from primed pluripotent stem cells. Although primed pluripotent stem cells are able to differentiate into tissue specific stem cells or differentiated cells, details in similarity of the global gene expression patterns are not better than naive pluripotent stem cell-derived somatic cells, if these different gene expression patterns might not affect functionality of the cells. hESCs have many features that are more similar to mouse EpiSCs than to mouse ESCs, including their morphology, the signaling pathways required for self-renewal, and their propensity for X chromosome inactivation (Hanna et al., 2010b; Silva et al., 2008). Moreover, EpiSCs have differentiation traits similar to both hESCs and human induced pluripotent stem cells (iPSCs) (Tesar et al., 2007). Interestingly, hESCs may bypass the tissue-specific stem cell state during in vitro differentiation. Koch et al. showed that although it is possible to derive hESC-derived NSCs that retain sufficient plasticity to be recruited into different neuronal subtypes in a controlled manner, the NSCs were not able to self-renew in long-term culture (passage 6–10) (Koch et al., 2009). In the present study, we established NSCs from EpiSCs that propagated more than 20 passages, which were also obtained in brain- and ESC-derived NSCs in the presence of EGF and bFGF (Conti et al., 2005). These results demonstrated that, unlike human primed pluripotent stem cells, mouse primed pluripotent stem cells could differentiate into long-term expandable NSCs in vitro. Recent reports have shown that primed pluripotent hESCs and hiPSCs can convert to the naive pluripotent state (Gu et al., 2012; Hanna et al., 2010a). However, whether naive pluripotent hESCs can differentiate into specific somatic stem cells has yet to be addressed. In our study, EpiSCs maintained their differentiation potential when cultured in an EpiSC medium supplemented with activin A (20 ng/ml) and bFGF (5 ng/ml). To derive NSCs, EpiSCs were cultured in high concentrations of bFGF (15 ng/ml) at the initial stage of differentiation. The differentiated cells formed neurospheres efficiently, which can be also established as NSC lines later on. It was previously shown that FGF2 signaling was inhibitory for neuroectodermal differentiation from EpiSCs and hESCs (Greber et al., 2010, 2011). The authors suggested that inhibition of FGF2 signaling rapidly downregulated NANOG and upregulated PAX6, leading to neural cell generation. In contrast, a recent study showed that sustained FGF signaling efficiently activated MAPK and induced expression of neural markers (PAX6, SOX2) in hESCs

515 (Lotz et al., 2013). Notably, FGF signaling is also essential for self-renewal and maintenance of NSCs (Conti et al., 2005; Xiao et al., 2007). In the present study, we showed that high concentrations of bFGF could trigger the induction of NSC in EpiSCs, which may provide useful clues to understanding of cell-fate transition into the neuronal lineage from the epiblast-stage embryos. Moreover, the mouse EpiSC-based system may provide a useful model system to study the development and differentiation of early post-implantation stage human embryos, which has not previously been feasible. Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.scr.2013.12.012.

Author contributions Hyo Jin Jang: collection and/or assembly of data, data analysis and interpretation, and manuscript writing Jong Soo Kim: collection and/or assembly of data, data analysis and interpretation, and manuscript writing Hyun Woo Choi: collection and/or assembly of data, data analysis and interpretation, and manuscript writing Sol Choi: data analysis and interpretation Min Jeong Kim: data analysis and interpretation Iksoo Jeon: data analysis and interpretation Jihwan Song: data analysis and interpretation and manuscript writing Jeong Tae Do: conception and design, data analysis and interpretation, manuscript writing, and final approval of manuscript

Acknowledgements This work was supported by the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (Grant 20110019489), and by the Next Generation BioGreen 21 Program funded by the Rural Development Administration (Grant PJ00956202).

References Aoki, H., Hara, A., Niwa, M., Yamada, Y., Kunisada, T., 2009. In vitro and in vivo differentiation of human embryonic stem cells into retina-like organs and comparison with that from mouse pluripotent epiblast stem cells. Dev. Dyn. 238, 2266–2279. Brons, I.G., Smithers, L.E., Trotter, M.W., Rugg-Gunn, P., Sun, B., de Sousa, Chuva, Lopes, S.M., et al., 2007. Derivation of pluripotent epiblast stem cells from mammalian embryos. Nature 448, 191–195. Chojnacki, A.K., Mak, G.K., Weiss, S., 2009. Identity crisis for adult periventricular neural stem cells: subventricular zone astrocytes, ependymal cells or both? Nat. Rev. Neurosci. 10, 153–163. Conti, L., Pollard, S.M., Gorba, T., Reitano, E., Toselli, M., Biella, G., et al., 2005. Niche-independent symmetrical self-renewal of a mammalian tissue stem cell. PLoS Biol. 3, e283. Deng, J., Steindler, D.A., Laywell, E.D., Petersen, B.E., 2003. Neural trans-differentiation potential of hepatic oval cells in the neonatal mouse brain. Exp. Neurol. 182, 373–382. Deng, J., Petersen, B.E., Steindler, D.A., Jorgensen, M.L., Laywell, E.D., 2006. Mesenchymal stem cells spontaneously express neural proteins in culture and are neurogenic after transplantation. Stem Cells 24, 1054–1064.

Gage, F.H., 2000. Mammalian neural stem cells. Science 287, 1433–1438. Gardner, R.L., Lyon, M.F., Evans, E.P., Burtenshaw, M.D., 1985. Clonal analysis of X-chromosome inactivation and the origin of the germ line in the mouse embryo. J. Embryol. Exp. Morphol. 88, 349–363. Gillich, A., Bao, S., Grabole, N., Hayashi, K., Trotter, M.W., Pasque, V., et al., 2012. Epiblast stem cell-based system reveals reprogramming synergy of germline factors. Cell Stem Cell 10, 425–439. Greber, B., Wu, G., Bernemann, C., Joo, J.Y., Han, D.W., Ko, K., et al., 2010. Conserved and divergent roles of FGF signaling in mouse epiblast stem cells and human embryonic stem cells. Cell Stem Cell 6, 215–226. Greber, B., Coulon, P., Zhang, M., Moritz, S., Frank, S., Muller-Molina, A.J., et al., 2011. FGF signalling inhibits neural induction in human embryonic stem cells. EMBO J. 30, 4874–4884. Gu, Q., Hao, J., Zhao, X.Y., Li, W., Liu, L., Wang, L., et al., 2012. Rapid conversion of human ESCs into mouse ESC-like pluripotent state by optimizing culture conditions. Protein Cell 3, 71–79. Guo, G., Yang, J., Nichols, J., Hall, J.S., Eyres, I., Mansfield, W., et al., 2009. Klf4 reverts developmentally programmed restriction of ground state pluripotency. Development 136, 1063–1069. Han, D.W., Tapia, N., Joo, J.Y., Greber, B., Arauzo-Bravo, M.J., Bernemann, C., et al., 2010. Epiblast stem cell subpopulations represent mouse embryos of distinct pregastrulation stages. Cell 143, 617–627. Han, D.W., Greber, B., Wu, G., Tapia, N., Arauzo-Bravo, M.J., Ko, K., et al., 2011. Direct reprogramming of fibroblasts into epiblast stem cells. Nat. Cell Biol. 13, 66–71. Hanna, J., Cheng, A.W., Saha, K., Kim, J., Lengner, C.J., Soldner, F., et al., 2010a. Human embryonic stem cells with biological and epigenetic characteristics similar to those of mouse ESCs. Proc. Natl. Acad. Sci. U. S. A. 107, 9222–9227. Hanna, J.H., Saha, K., Jaenisch, R., 2010b. Pluripotency and cellular reprogramming: facts, hypotheses, unresolved issues. Cell 143, 508–525. Huang da, W., Sherman, B.T., Lempicki, R.A., 2009. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protoc. 4, 44–57. Iwashita, Y., Crang, A.J., Blakemore, W.F., 2000. Redistribution of bisbenzimide Hoechst 33342 from transplanted cells to host cells. Neuroreport 11, 1013–1016. Jouneau, A., Ciaudo, C., Sismeiro, O., Brochard, V., Jouneau, L., Vandormael-Pournin, S., et al., 2012. Naive and primed murine pluripotent stem cells have distinct miRNA expression profiles. RNA 18, 253–264. Koch, P., Opitz, T., Steinbeck, J.A., Ladewig, J., Brustle, O., 2009. A rosette-type, self-renewing human ES cell-derived neural stem cell with potential for in vitro instruction and synaptic integration. Proc. Natl. Acad. Sci. U. S. A. 106, 3225–3230. Lotz, S., Goderie, S., Tokas, N., Hirsch, S.E., Ahmad, F., Corneo, B., et al., 2013. Sustained levels of FGF2 maintain undifferentiated stem cell cultures with biweekly feeding. PLoS One 8, e56289. Meissner, A., Mikkelsen, T.S., Gu, H., Wernig, M., Hanna, J., Sivachenko, A., et al., 2008. Genome-scale DNA methylation maps of pluripotent and differentiated cells. Nature 454, 766–770. Najm, F.J., Chenoweth, J.G., Anderson, P.D., Nadeau, J.H., Redline, R.W., McKay, R.D., et al., 2011a. Isolation of epiblast stem cells from preimplantation mouse embryos. Cell Stem Cell 8, 318–325. Najm, F.J., Zaremba, A., Caprariello, A.V., Nayak, S., Freundt, E.C., Scacheri, P.C., et al., 2011b. Rapid and robust generation of functional oligodendrocyte progenitor cells from epiblast stem cells. Nat. Methods 8, 957–962. Nichols, J., Smith, A., 2009. Naive and primed pluripotent states. Cell Stem Cell 4, 487–492. Silva, S.S., Rowntree, R.K., Mekhoubad, S., Lee, J.T., 2008. Xchromosome inactivation and epigenetic fluidity in human

516 embryonic stem cells. Proc. Natl. Acad. Sci. U. S. A. 105, 4820–4825. Stadtfeld, M., Hochedlinger, K., 2010. Induced pluripotency: history, mechanisms, and applications. Genes Dev. 24, 2239–2263. Temple, S., 2001. The development of neural stem cells. Nature 414, 112–117. Tesar, P.J., Chenoweth, J.G., Brook, F.A., Davies, T.J., Evans, E.P., Mack, D.L., et al., 2007. New cell lines from mouse epiblast share defining features with human embryonic stem cells. Nature 448, 196–199.

H.J. Jang et al. Vallier, L., Touboul, T., Chng, Z., Brimpari, M., Hannan, N., Millan, E., et al., 2009. Early cell fate decisions of human embryonic stem cells and mouse epiblast stem cells are controlled by the same signalling pathways. PLoS One 4, e6082. Xiao, Z., Kong, Y., Yang, S., Li, M., Wen, J., Li, L., 2007. Upregulation of Flk-1 by bFGF via the ERK pathway is essential for VEGF-mediated promotion of neural stem cell proliferation. Cell Res. 17, 73–79. Ying, Q.L., Wray, J., Nichols, J., Batlle-Morera, L., Doble, B., Woodgett, J., et al., 2008. The ground state of embryonic stem cell self-renewal. Nature 453, 519–523.