Rapid induction and long-term self-renewal ofprimitive neural precursors from human embryonicstem cells by small molecule inhibitorsWenlin Lia,b, Woong Sunc,d, Yu Zhanga, Wanguo Weia, Rajesh Ambasudhana, Peng Xiae, Maria Talantovae,Tongxiang Lina, Janghwan Kima, Xiaolei Wangc, Woon Ryoung Kimd, Stuart A. Liptone, Kang Zhangc,f,1,and Sheng Dinga,g,1

aDepartment of Chemistry, The Scripps Research Institute, La Jolla, CA 92037; bDepartment of Cell Biology, Second Military Medical University, Shanghai200433, China; cInstitute for Genomic Medicine and Shiley Eye Center, University of California, San Diego, CA 92093; dDepartment of Anatomy, KoreaUniversity College of Medicine, Brain Korea 21 Program, Seoul, 136-705, Korea; eDel E. Webb Center for Neuroscience, Aging, and Stem Cell Research,Sanford-Burnham Medical Research Institute, La Jolla, CA 92037; fMolecular Medicine Research Center and Department of Ophthalmology, West ChinaHospital, Sichuan University, Chengdu 610065, China; and gGladstone Institute of Cardiovascular Disease, Department of Pharmaceutical Chemistry,University of California, San Francisco, CA 94158

Edited by Fred H. Gage, The Salk Institute, San Diego, CA, and approved March 28, 2011 (received for review September 20, 2010)

Human embryonic stem cells (hESCs) hold enormous promise forregenerative medicine. Typically, hESC-based applications wouldrequire their in vitro differentiation into a desirable homogenouscell population. A major challenge of the current hESC differenti-ation paradigm is the inability to effectively capture and, in thelong-term, stably expand primitive lineage-specific stem/precursorcells that retain broad differentiation potential and, more impor-tantly, developmental stage-specific differentiation propensity.Here, we report synergistic inhibition of glycogen synthase kinase3 (GSK3), transforming growth factor β (TGF-β), and Notch signal-ing pathways by small molecules can efficiently convert mono-layer cultured hESCs into homogenous primitive neuroepitheliumwithin 1 wk under chemically defined condition. These primitiveneuroepithelia can stably self-renew in the presence of leukemiainhibitory factor, GSK3 inhibitor (CHIR99021), and TGF-β receptorinhibitor (SB431542); retain high neurogenic potential and respon-siveness to instructive neural patterning cues toward midbrainand hindbrain neuronal subtypes; and exhibit in vivo integration.Our work uniformly captures and maintains primitive neural stemcells from hESCs.

Human embryonic stem cells (hESCs) hold enormous promisefor regenerative medicine (1). Typically, hESC-based appli-cations require in vitro differentiation of hESCs into a desirablehomogenous cell population. Despite the enormous progressesmade in differentiating hESCs into various functional cells, amajor challenge of the current hESC differentiation paradigm isthe inability to effectively capture and stably expand primitivelineage-specific stem/precursor cells. These cells would ideallyretain broad differentiation potentials (e.g., have the ability toserially repopulate the entire specific tissue) and, perhaps moreimportantly, the developmental stage-specific differentiationpropensity, and would be devoid of tumorigenicity concerns. Inthe case of neural induction of hESCs by various advancedmethods (2–5), there is still a lack of robust, chemically definedconditions for the long-term maintenance of primitive neuralepithelial precursor cells, which are highly neurogenic and can bepatterned/regionalized by specific morphogens (6, 7). Undertypically used growth factor conditions (including bFGF, EGF),neural stem cells (NSCs) “transition” in a few passages into amore glial-restricted precursor state (8), which is significantly lessneurogenic. In addition, in vitro cultured NSCs respond poorlyto patterning cues and exhibit a narrow repertoire for generatingspecific neuronal subtypes. Previous studies in murine ESCs(mESCs) have suggested the existence of leukemia inhibitoryfactor (LIF)-responsive primitive NSCs (6). However, these cellscould not be maintained in culture. Recent studies in neural in-duction of hESCs have identified rosette-type NSCs that repre-sent neural tube-stage precursor cells. These rosette NSCs werecapable of responding to patterning cues that direct differentia-

tion toward region-specific neuronal fates, but still could not bestably maintained (4). Recently, Koch et al. reported long-termexpansion of hESC-derived rosette-type NSCs (9). However, thestudy used the conventional and undefined embryoid body (EB)differentiation strategy and required tedious mechanical isolationof the overgrown neural rosettes from replated EBs. In addition,under these conditions, NSCs could not maintain stable spatialproperties and switch from forebrain to hindbrain identity afterprolonged expansion.In our attempts to convert conventional hESCs to a mESC-like

naïve state by small molecules, we fortuitously created a homoge-nously converted cell population by combined treatment of humanLIF (hLIF) and two small molecules, CHIR99021 and SB431542,for about 10 d under chemically defined conditions. Remarkably,this population of cells, growing in colonies, appeared to self-renewand stably maintain their characteristics over numerous passagesunder these defined conditions. CHIR99021 (referred to hereafteras CHIR) is a small molecule inhibitor of glycogen synthase kinase3 (GSK3) and can activate canonical Wnt signaling (10), whichhas been implicated in ES cell self-renewal (11). SB431542 (re-ferred to hereafter as SB) is a small molecule inhibitor of trans-forming growth factor β (TGF-β) and Activin receptors, and hasbeen implicated in the mesenchymal-to-epithelial transition andreprogramming (12, 13). Interestingly, these converted cells didnot express the pluripotency markers Oct4 and Nanog, but werepositive for Sox2 and alkaline phosphatase (ALP). Subsequentstudies revealed that this expandable cell population has featuresof primitive neuroepithelium (and hereafter we refer them asprimitive neural stem cells/pNSCs). Interestingly, the self-renewalof pNSCs is dependent on LIF, which has been implicated in theself-renewal of mESC-derived primitive NSCs (6, 14). Previous invivo developmental studies have shown that bFGF-responsivedefinitive NSCs first appear on embryonic day 8.5 (ED 8.5) inmouse embryos (15, 16).However, at an earlier stage (ED 5.5–7.5),primitive NSCs are LIF-dependent, and the in vivo generation ofprimitive NSCs was independent of Notch signaling. We reasonedthat if pNSCs are analogous to primitive NSCs during de-

Author contributions: W.L., K.Z., and S.D. designed research; W.L., W.S., Y.Z., W.W., R.A.,P.X., M.T., T.L., X.W., andW.R.K. performed research; W.W., T.L., and J.K. contributed newreagents/analytic tools; and W.L., W.S., R.A., P.X., S.A.L., K.Z., and S.D. analyzed data; andW.L. and S.D. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Data deposition: The data reported in this paper have been deposited in the Gene Ex-pression Omnibus (GEO) database, www.ncbi.nlm.nih.gov/geo (accession no. GSE28595).1To whom correspondence may be addressed. E-mail: k5zhang@ucsd.edu or sheng.ding@gladstone.ucsf.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1014041108 PNAS | May 17, 2011 | vol. 108 | no. 20 | 8299–8304

DEV

ELOPM

ENTA

LBIOLO

GY

Dow

nloa

ded

by g

uest

on

June

15,

202

1

www.ncbi.nlm.nih.gov/geohttp://www.pnas.org/external-ref?link_type=NCBIGEO&access_num=GSE28595mailto:k5zhang@ucsd.edumailto:sheng.ding@gladstone.ucsf.edumailto:sheng.ding@gladstone.ucsf.eduhttp://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplementalhttp://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplementalwww.pnas.org/cgi/doi/10.1073/pnas.1014041108

velopment, temporarily inhibitingNotch signaling should not blockthe induction of pNSCs. Indeed, temporal treatment by anothersmall molecule inhibitor of γ-secretase, Compound E (referred tohereafter as C-E; ref. 17), further accelerated neural induction andgenerated the homogenous self-renewing pNSC population within1wk.Even after long-termexpansion and repeated passaging in thepresence of hLIF, CHIR, and SB, pNSCs retain remarkably highneurogenic propensity, broad differentiation potential, respon-siveness to extrinsic morphogens for subsequent development intosubtype-specific neuronal identities, and the ability to integrate invivo. Our work uniformly captures and stably maintains primitiveneural stem cells from hESCs.

ResultsSynergistic Inhibition of GSK3, TGF-β, and Notch Signaling PathwaysConverts hESCs into Homogenous pNSCs. hESCs were cultured onX-ray inactivated CF-1 mouse embryonic fibroblasts (MEFs) inhESC growth media (DMEM/F12 containing 20% KSR and 10ng/mL bFGF) or on Matrigel under feeder-free and chemicallydefined conditions as described (18). Primitive neuroepitheliumwas induced by switching from hESC growth media to neural in-duction media (1:1 Advanced DMEM/F-12:Neurobasal mediasupplemented with N2, B27, hLIF), supplemented with CHIRand SB, with or without C-E, for 7 d (a schematic representationof the differentiation process is shown in Fig. 1A). hESC differ-entiation wasmonitored by immunocytochemistry, flow cytometryand real-time PCR. As shown in images of the same visual field,the differentiated cells exhibited homogenous epithelial mor-phology during the entire differentiation process (Fig. S1A 1–4).Real-time PCR analysis revealed that combined treatment withhLIF, SB, and CHIR (with or without C-E) induced a rapid loss ofOct4 and Nanog expression (Fig. 1B). However, the expressionof Sox2, a pluripotency marker that is also a persistent marker ofNSCs (19), remained largely unchanged. Pax6, an early markerof neural induction, was significantly up-regulated after 5 d inthe presence of C-E (0.1 μM), whereas its up-regulation was firstdetected at the sixth day in the absence of C-E treatment (Fig.1B). Consistent with this observation, immunocytochemistryconfirmed the faster induction of Pax6 protein on the sixth day inthe presence of C-E, as Pax6 protein only became detectable fromthe seventh day onwards in the absence of C-E (Fig. S1B 7–12). Incontrast, only a small fraction of cells were positive for Pax6 onday 7 when hESCs were treated with SB, C-E, and hLIF (Fig.S1C1). Similarly, no Pax6 positive cells could be detected at thesame time point when hESCs were treated with CHIR, C-E, andhLIF (Fig. S1C2). These data suggest that inhibition of Notchsignaling can enhance early neural induction. Interestingly, real-time PCR analysis showed that the induction of the Pax6 geneoccurred in parallel with the suppression of BMP4 gene expres-sion as well as induction of Noggin (BMP antagonist) expression(Fig. 1B), suggesting that endogenous mechanisms of BMP sig-naling inhibition may contribute to neural induction. Real-timePCR analysis also demonstrated that the differentiation is highlyspecific toward the neural lineage. Along with the induction ofPax6, epiblast-associated nonneural genes such as Brachyury,Eomes, and Sox17, were repressed synchronously with pluri-potency markers Oct4 and Nanog (Fig. 1B), suggesting the pres-ence of an intermediate cell type resembling differentiatingepiblast cells before hESC neuralization. This highly directedneural induction was further confirmed by immunocytochemistry.Double staining of Oct4 and Nestin showed that Oct4 expressiongradually diminished and was almost undetectable after 5 d oftreatment with hLIF, SB, CHIR, and C-E, whereas Nestin-expressing cells became the predominant population, comprising∼99% of the population on day 7 (Fig. S1B 1–6). To furtherquantify the efficiency of the neural induction, the expression ofOct4, Sox2 and CD133, was analyzed by flow cytometry. In de-velopment, the neural plate and neural tube exhibit CD133(Prominin-1) immunoreactivity (20, 21). In vertebrate embryos,Sox2 is one of the earliest markers for the neural plate. DuringhESC differentiation, the earliest Oct4-negative, but Sox2/

CD133-positive cell population would represent the primitiveneuroepithelium. FACS analysis showed that more than 96% ofundifferentiated hESCs were positive for both Oct4 and Sox2(Fig. 1C). After treatment, FACS confirmed the rapid loss of Oct4expression. Especially Oct4-positive cell number dropped sub-stantially on day 5, when Pax6 was first induced, suggesting thatday 5 was the turning point of neural induction. In addition, FACSanalysis further showed that the addition of C-E induced a muchmore rapid loss of Oct4 expression and consequent neural con-version. At day 5, only 13% of cells were still positive for Oct4 inthe presence of C-E, whereas 33.9% were positive in its absence.Despite the loss of Oct4 expression, cells persistently maintaineda high level of Sox2 expression (>96%) at all time points exam-ined during differentiation, and >97% of cells were only positivefor Sox2 at day 7 with C-E treatment (Fig. 1C). In addition, FACSanalysis showed that 98%of undifferentiated hESCs were positivefor CD133 and that small molecule treatment initially induced theloss of CD133. However, along with the induction of Pax6 fromday 5 onwards, the CD133-positive cells increased significantly,

Fig. 1. Real-time PCR and flow cytometry analysis of neural induction fromhESCs treated with LIF, CHIR, and SB (with or without C-E). (A) Schematicrepresentation of the neural induction process. (B) The expression of Pax6,Sox2, Nanog, Oct4, BMP4, Noggin, Eomes, Brachyury, and Sox17 was ana-lyzed by real-time PCR. (C and D) Flow cytometry analysis was used toquantify cells expressing Oct4, Sox2, or CD133 during neural induction. CDM,chemically defined medium.

8300 | www.pnas.org/cgi/doi/10.1073/pnas.1014041108 Li et al.

Dow

nloa

ded

by g

uest

on

June

15,

202

1

http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF1www.pnas.org/cgi/doi/10.1073/pnas.1014041108

with >98% cells being CD133-positive on day 9 (Fig. 1D). Thesehomogenously differentiated neural cells could be stably ex-panded on MEF feeder cells or Matrigel coating in the presenceof hLIF, CHIR, and SB, and are referred to hereafter as pNSCs.In the present study, pNSCs were regularly expanded onMatrigel.Taken together, these data suggested that the combination ofhLIF/CHIR/SB/C-E directs the specific induction of primitiveneuroepitheliumwithin 7 d that can long-term homogenously self-renew under hLIF/CHIR/SB conditions without the need for anycell purification. Chambers et al. (2) recently demonstrated thatdual inhibition of Smad signals by Noggin and SB431542 couldconvert>80% hESCs to neural fate in 13 d. However, the Noggin/SB431542 condition (which also contains undefined serumproducts) generated heterogeneous neural populations contain-ing cells of different developmental stages (e.g., nonpolarizedneuroepithelia and polarized rosette-like structures). Most im-portantly, the dual Smad inhibition protocol cannot capture theNSCs and maintain their self-renewal. Our described neural in-duction process is much faster, more specific, and more efficient,representing a chemically defined single-step strategy for ob-taining self-renewing homogenous primitive NSCs from hESCscultured in a monolayer. The results of our strategy are highlyreproducible in multiple different hESC lines, including H1 (Fig.1), HUES9, and HUES1 (Fig. S2A), under both feeder andfeeder-free (Matrigel) culture conditions.

pNSCs Can Long-Term Self-Renew and Represent the Pre-RosetteStage NSCs. pNSCs can long-term self-renew over serial pas-sages on Matrigel with SB, CHIR, and hLIF. pNSCs generatedfrom HUES9 and H1 hESCs were routinely passaged 1:10 andhave been cultured for >30 passages without obviously losingproliferative capacity, which is equivalent to at least 94 pop-ulation doublings. However, individual omission of hLIF, SB, orCHIR from the media compromised pNSCs’ long-term self-renewal. Single pNSC is clonogenic on Matrigel in the presenceof hLIF/SB/CHIR (Fig. 2A). However, no colonies were observedunder conditions including C-E, suggesting that Notch signalingis critical to pNSC self-renewal. Consistently, treatment with C-Efor 48 h rapidly induced pNSCs to differentiate into Doublecortin(DCX)-positive neuronal precursors (Fig. S2B 1 and 2). Despitetheir highly proliferative and clonogenic capacity, pNSCs are nottumorigenic in SCID beige mice. We transplanted the early-passage (passage 6, about 30 d in serial culture) and late-passage(about passage 27) of HUES9- and H1-derived pNSCs (2 × 106cells suspended in Matrigel) into 24 SCID beige mice s.c. Thesemice have been observed for as long as 6 mo with no sign ofneoplasm formation, whereas the control animals transplantedwith the parental hESCs produced teratomas within 6 wk.Remarkably, the long-term expanded pNSCs maintain a stable

pNSC phenotype. The pNSCs cultured on Matrigel exhibitedtypical epithelial morphology and positive ALP staining (Fig. 2A).Immunostaining showed that both the early-passage (passage 6)and late-passage (passage 27) pNSCs stably expressed genes re-cently identified as rosette-type NSCmarkers (4), including PLZF(promyelocytic leukemia zinc finger), ZO-1, and N-cad (N-cad-herin); CNS (central nervous system) neural stem cell markers,such as Nestin, Pax6, and Sox2; anterior neural markers Forse1and Otx2; and the midbrain marker Nurr1 (Fig. 2 B–I and Fig.S3A 1–8). Expression analysis by microarray confirmed the dra-matic up-regulation of neural lineage genes such as Ascl1, Pax6,Dach1, N-cad, and Nestin, and down-regulation of pluripotencygene Oct4 in pNSCs in comparision to hESCs. However, bothhESCs and pNSCs express ZO-1 and Sox2 at similar level. Evenafter long-term passaging, pNSCs uniformly expressed a panel ofprimitive neuroepithelial genes, including Sox2, N-cad, PLAZ,Dach1, ZO-1, Pax6, and proneuronal gene Ascl1, and both early-and late-passage pNSCs demonstrated highly similar tran-scriptome profile (Fig. S3B 1 and 2). Notably, N-cad and the tightjunction protein ZO-1 were expressed evenly on the surface ofboth early- and late-passage pNSCs, suggesting that pNSCs areprimitive, nonpolarized prerosette NSCs (2). Indeed, pNSCs

gained rosette-like structures with apical N-cad expression andinterkinetic nuclear migration after being cultured in neural in-duction media with 20 ng/mL bFGF for 4 d (Fig. S3C 1 and 2).Consistent with their highly proliferative capacity, pNSCs uni-formly expressed Ki-67 (Fig. 2F and Fig. S3A5). The stable phe-notype of pNSCs after extensive passaging was further confirmedby flow cytometry. Both early-passage and late-passage pNSCsexhibited nearly identical expression patterns for a set of NSC-specific markers, such as Nestin (98.2% positive), Pax6 (95.4%positive), CD133 (93.9% positive), and the cell proliferationmarker Ki-67 (98.3% positive; Fig. 2J), whose uniform expressionconfirmed that pNSCs were a homogenous, expandable NSCpopulation. Indeed, genes associatedwith non-neural lineages, suchas Eomes, Brachyury, Sox17, or K15 were undetectable in pNSCsby RT-PCR (Fig. 2K). FACS analysis with propidium iodiderevealed a very similar cell cycle profile for both early- and late-passage of pNSCs. The cell cycle distribution (G1, S, and G2/M) ofearly-passage pNSCs (P7) is 52.7%, 26.6%, and 16.5%; and thecycle distribution of late-passage pNSCs (P28) is 51.7%, 32.4%,and 12.4%, respectively. Interestingly, FACS analysis showed thatthe expression of Forse1, an anterior NSC marker, was not ho-mogenous (53.6% of pNSCs were positive for Forse1; Fig. 2J).

Fig. 2. pNSCs stably self-renew and maintain a homogenous primitive NSCphenotype after long-term cultures. The pNSCs cultured on Matrigelexhibited characteristic epithelial morphology. (A) A single cell-derived pNSCcolony on Matrigel. (Inset) pNSCs were positive for ALP. (B–I) Immunocyto-chemistry showed that pNSCs (passage 6) expressed genes recently identifiedas rosette-type NSC markers, including PLZF, ZO-1, and N-cad; CNS neuralstem cell makers such as Nestin, Pax6, and Sox2; the cell proliferation markerKi-67; the anterior neural markers Forse1 and Otx2; and the midbrain markerNurr1. (J) Flow cytometry analysis showed that pNSCs stably expressed NSCand cell proliferation markers after long-term in vitro expansion, includingNestin, Pax6, CD133, Forse1, and Ki-67. (K and L) Nonneural lineage markersand genes associated with midbrain were analyzed by RT-PCR.

Li et al. PNAS | May 17, 2011 | vol. 108 | no. 20 | 8301

DEV

ELOPM

ENTA

LBIOLO

GY

Dow

nloa

ded

by g

uest

on

June

15,

202

1

http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF2http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF2http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF3http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF3

Whether Forse1-negative pNSCs have a more posterior identityneeds to be further characterized. In addition, pNSCs did notexpress neural crest cell markers such as HNK1, Sox10, or p75.However, we did detect a small percentage (∼3%) of pNSCspositive for AP2, a premigratory neural crest gene initiallyexpressed throughout the neural plate border (22). To rule out thepossibility that pNSC cultures contained a separate (parallel orunrelated) neural crest cell population, we preformed clonalanalysis by immunostaining of single pNSC-derived colonies.AP2-positive cells representing 2–5% of cells in each colony weredetected in all examined colonies (n = 16), and they were mostlyseen at the border of the colonies (Fig. S4), suggesting that theywere derivatives of pNSCs. However, whether these AP2-positivecells possess neural crest potential remains to be confirmed.To examine the multipotency of long-term expanded pNSCs,

both early- and late-passage pNSCs were plated at ultra-lowdensity in six-well plates (200 cells per well) and cultured in dif-ferentiation media for 2 wk. Among the single pNSC-derived cellclusters (n = 29), 100% contained both MAP2-positive neuronsand GFAP-positive astrocytes (Fig. S5 A–C), but no cells positivefor the oligodendrocyte marker O4 or the neural crest lineagemarkers peripherin and α-SMA were detected at this time point.Previous studies have shown that bFGF and/or EGF-expandedNSCs lose neurogenic propensity and become more gliogenicafter long-term culture (23). However, pNSCs expanded underour described conditions retained high neuronal differentiationpropensity. Flow cytometry analysis showed that pNSCs at pas-sage 8 and passage 25 could give rise 73.9% and 77.6% MAP2-positive, or 71.4% and 74.4% NeuN-positive neurons, respecti-vely (Fig. 3A 1 and 2). During CNS development, neurogenesislargely precedes gliogenesis. NSCs from earlier stages generatemore neurons and have a lower propensity to produce glia thanthose from later stages. The remarkably high neurogenic poten-tial and propensity of these long-term expanded pNSCs is con-sistent with their self-renewal in the primitive state. Importantly,pNSCs could effectively differentiate and generate mature neu-rons that fired action potentials (5 cells in 7 tested cells; Fig. 3B),and produced fast inactivating inward Na+ currents (n = 8 of8 cells recorded) that were sensitive to the Na+ channel blockerTetrodotoxin (TTX; Fig. 3C). Furthermore, these differentiatedneurons manifested spontaneous excitatory postsynaptic currents(sEPSCs) and/or inhibitory postsynaptic currents (sIPSCs) in 4 of6 cells recorded (Fig. 3D), indicating that they can form func-

tional synapses. Next, to further examine pNSCs’ potential invivo, they were transplanted into the lateral ventricle of neonatalmice (P2-3). Histological analysis of GFP-expressing pNSC(passage 27) grafts one month after transplantation revealed thatengrafted cells were distributed in many brain areas, including thecorpus callosum, the subcallosal zone, the caudate-putamen (Fig.S6 A and B), and the hindbrain (Fig. S6 C–J). Most engrafted cells(>50% in the forebrain, and >80% in the hindbrain) express-ed differentiated neuronal markers such as MAP2 (Fig. S6A 1–4).In addition, we also detected DCX-positive engrafted cells in thesubcallosal zone (Fig. S6B 1–4), where endogenous adult neuralprogenitor cells reside (24), but not in non-neurogenic environ-ments such as the hindbrain, suggesting that their neuronal dif-ferentiation was influenced by the host environment. Althoughwe failed to detect the mature neuronal marker NeuN in thesubcallosal zone or caudate-putamen, a subset of GFP-express-ing cells in the clusters near the aqueduct exhibited NeuN ex-pression (Fig. S6C 1–4). We also failed to detect spontaneouslydifferentiated tyrosine hydroxylase (TH)-positive dopaminergic(DA) neurons, but some engrafted cells (∼10%) appeared tohave differentiated into GABA-expressing inhibitory neurons(Fig. S6D 1–4). GFP-expressing cells in the hindbrain, closelyassociated with presynaptic puncta labeled by synaptophysin,were also observed, indicating the synaptic contacts of thetransplanted cells with the host mouse neurons (Fig. S6E 1–4).All GFP positive cells also exhibited human nucleus antigenimmunoreactivity (Fig. S6F 1–4), further confirming their humancell identity. In addition, some GFP-expressing hindbrain neu-rons also exhibited c-fos, a marker for neuronal excitation (Fig.S6 G–I). On the other hand, we did not find any GFAP-positiveengrafted cells (Fig. S6J), suggesting that pNSCs preferentiallydifferentiate into the neuronal lineage in vivo.

pNSCs Possess Mesencephalic Regional Identity and Can Be Re-specified Toward Caudal Cell Fates. It is worthwhile to note thatpNSCs express the forebrain/midbrain gene Otx2 and the mid-brain gene Nurr1 by immunostaining (Fig. 2 H and I). RT-PCRanalysis confirmed the expression of Otx2 and Nurr1, and showedthat pNSCs also express other midbrain genes, such as En-1,Lmx1b, Pax2, and Pitx3 (Fig. 2L). In contrast, the forebrain-re-stricted transcription factors FoxG1 and Emx2 were barely de-tectable, and anterior hindbrain transcription factors, such asGbx2, HoxB2, and HoxA2, were expressed at low levels as in-

Fig. 3. pNSCs retain high neurogenic potential during long-term culture. (A) Flow cytometry analysis showed pNSCs atpassage 8 and passage 25 could give rise to 73.9% and 77.6%MAP2-positive, or 71.4% and 74.4% NeuN-positive neurons,respectively. (B) Representative traces of evoked actionpotentials (whole-cell recording, current-clamp mode) gener-ated by neurons after 4 wk of differentiation from pNSCs.Traces of Tetrodotoxin (TTX)-sensitive whole-cell currentsrecorded in voltage-clamp mode. (C) Cells were hyperpolarizedto −90 mV for 300 ms before applying depolarizing pulses toelicit Na+ and K+ currents. (D) Traces of spontaneous excitatorypostsynaptic currents (sEPSCs) and spontaneous inhibitorypostsynaptic currents (sIPSCs), both recorded at a holding po-tential of −60 mV, indicated synapse formation. B–D representthe data recorded from pNSCs at passage 25 that had spon-taneously differentiated to display neuronal properties.

8302 | www.pnas.org/cgi/doi/10.1073/pnas.1014041108 Li et al.

Dow

nloa

ded

by g

uest

on

June

15,

202

1

http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF4http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF5http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF6www.pnas.org/cgi/doi/10.1073/pnas.1014041108

dicated by RT-PCR (Fig. S7A). pNSCs expressed the dorsalneural tube gene Pax3, the ventral neural tube genes Nkx2.2 andNkx6.1, the NSC marker Dach1 and the Notch effector Hes5 (Fig.S7A). These observations suggested that in vitro-expandedpNSCs may possess a mesencephalic regional identity. To con-firm this, two different differentiation protocols were used toexamine the potential of pNSCs to generate midbrain DA neu-rons: in the “induced” protocol, pNSCs were first treated with 100ng/mL SHH (Sonic Hedgehog) and 100 ng/mL FGF8b for 10d and were then further differentiated in the presence of 10 ng/mL BDNF, 10 ng/mL GDNF, 10 ng/mL IGF1, 1 ng/mL TGF-β3,and 0.5 mM dibutyryl-cAMP (dbcAMP) for another ∼2-3 wk(Fig. 4A). In the “default” protocol, pNSCs were directly termi-nally differentiated in the presence of BDNF, GDNF, IGF1,TGF-β3, and dbcAMP for 3 wk without pre-patterning bymorphogens (Fig. 4B). Both the “induced” and “default” differ-entiation conditions produced >50% TH-positive neurons thatalso exhibited aromatic L-amino acid decarboxylase (AADC),En-1, Lmx1a, Nurr1, FoxA2, and Pitx3 immunoreactivity (Fig. 4A 1–7 and B 1–7). Notably, Pitx3 is a homeobox gene uniquelyexpressed in midbrain DA neurons (25). Real-time PCR con-firmed the significant up-regulation of TH, AADC, En-1, Nurr1,and Pitx3 (Fig. 4 A8 and B8). Flow cytometry quantificationdemonstrated that the “induced” and “default” differentiationprotocols produced 54.8% and 62.7% TH and MAP2 doublepositive neurons, respectively (Fig. 4 A9 and B9). These datademonstrate that pNSCs possess mesencephalic regional identityand can differentiate into DA neurons with very high efficiency.Because they exhibit features of pre-rosette primitive NSCs,

pNSCs were further examined for their responsiveness to in-structive regional patterning cues. pNSCs were sequentially trea-ted with caudalizing retinoic acid (RA, 1 μM) for 7 d, 100 ng/mLSHH and 0.1 μMRA for another 7 d, and then 50 ng/mL SHH and0.1 μM RA for an additional 7 d. The cells were then terminallydifferentiated in the presence of 10 ng/mL BDNF and 10 ng/mLGDNF in differentiation media for about 7 d. Real-time PCRassays demonstrated significant induction of posterior genes, in-cludingHoxB4,HoxA5, andHoxC5 after treatment with 1 μMRAfor 1 wk (Fig. S8A), suggesting that pNSCs are responsive to thecaudalizing effect of RA. Under such conditions, immunocyto-chemistry showed that pNSCs could differentiate into cholineacetyltransferase (ChAT)-positive neurons that are also positivefor MAP2 and Isl-1 (Fig. S8 B and C). Flow cytometry analysisshowed that 53.7% cells were double-positive for Isl-1 and MAP2(Fig. S8D). Real-time PCR assays confirmed the significant in-duction of ChAT, HB9, Isl-1, and Lim3 after terminal differenti-ation (Fig. S8A), suggesting an induction of motor neurons. Thesedata indicated that pNSCs retain responsiveness to instructivecues promoting the induction of hindbrain neuronal subtypes.

DiscussionTo realize the potential of cell-based therapy for treating injuriesand degenerative diseases, renewable sources of stem/progenitorcells need to be developed. Although hESCs indefinitely self-renew and have the differentiation potential to become any celltype, they are practically inferior to lineage-restricted cells as theyare prone to causing teratomas and do not repopulate host tissuesin vivo. However, significant challenges also remain in terms ofthe isolation and long-term expansion of most tissue-specificstem/progenitor cells from adults (e.g., even for the arguably moststudied hematopoietic stem cells). Consequently, differentiationof hESCs into renewable tissue-specific cell types is highly desir-able for various biomedical applications. If achieved, cell pop-ulations could be carefully quality controlled and serve as startingmaterials, skipping hESCs that cannot be used directly. Further-more, despite significant advances in development of variousneural induction conditions for hESCs, most differentiationprotocols use poorly defined culture conditions (e.g., goingthrough EB formation, using undefined medium supplements/KSR), and usually yield mixed populations containing neural cellsat different developmental stages, or even other embryonic germ

layer lineages and undifferentiated hESCs. In the present study,our serendipitous observation led us to develop a robust chemi-cally defined condition using specific small molecules that rapidlyand uniformly converts hESCs into pNSCs, and, most impor-tantly, enables their long-term expansion without a loss of highneurogenic propensity and regionalizable plasticity. To ourknowledge, this is the fastest and most efficient method so far toproduce neural stem cells from hESCs. In addition, pNSCs differfrom previously reported hESC-derived NSCs in that they rep-resent the primitive pre-rosette neuroepithelium that has neverbeen long-term expanded in vitro before. Interestingly, pNSCspossess features of mesencephalic precursor cells and can dif-ferentiate into DA neurons spontaneously with high efficiency inthe absence of pre-patterning. Real-time PCR analysis showedthe up-regulation of endogenous SHH, FGF8, and the ventralpatterning gene Nkx6.1 under both “induced” and “default” dif-ferentiation protocols (Fig. S7B), suggesting the cells could bespecified into DA neurons by an endogenous mechanism. Theseobservations are reminiscent of the previous in vivo studies thatshowed DA neurons originated from SHH-expressing domains ofthe ventral midbrain (26). In addition, a mouse study demon-

Fig. 4. pNSCs possess mesencephalic regional identity and can differentiateinto DA neurons with high efficency. (A) In the “induced” protocol, pNSCswere treated with SHH and FGF8b for 10 d before they were terminallydifferentiated in the presence of BDNF, GDNF, IGF1, TGF-β3, and dbcAMP foranother ∼2–3 wk. (B) In the “default” protocol, pNSCs were directly termi-nally differentiated in the presence of BDNF, GDNF, IGF1, TGF-β3, anddbcAMP for 3 wk. Under both protocols, pNSCs gave rise to TH positiveneurons that also exhibited AADC, En-1, Lmx1a, Nurr1, FoxA2, and Pitx3immunoreactivity (A 1–7 and B 1–7). Real-time PCR further confirmed thesignificant up-regulation of TH, AADC, En-1, Nurr1, and Pitx3 (A8 and B8).Flow cytometry quantification demonstrated that the two differentiationprotocols produced 54.8% and 62.7% TH and MAP2 double-positive neu-rons, respectively (A9 and B9).

Li et al. PNAS | May 17, 2011 | vol. 108 | no. 20 | 8303

DEV

ELOPM

ENTA

LBIOLO

GY

Dow

nloa

ded

by g

uest

on

June

15,

202

1

http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF7http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF7http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF7http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF8http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF8http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF8http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF8http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=SF7

strated the antagonistic interaction between the activation ofWnt/β-catenin and SHH (27). The activation of β-catenin in theventral midbrain promoted the expansion of early DA progeni-tors, but led to a reduced expression of SHH. The removal of theGSK3 inhibitor (CHIR) during pNSC differentiation may lead todown-regulation of Wnt/β-catenin signaling and facilitate the up-regulated SHH expression in turn. Considering the significance ofdeveloping renewable sources of DA neurons, it would be usefulto examine whether pNSC transplantation could attenuate theParkinson’s symptoms in animal models in the future.Recent studies suggest that GSK3 plays key roles in many

fundamental processes, including mediating signaling down-stream of Wnt, FGF, Hh, and Notch during neural development(28–30). In our neural induction protocol, however, replacementof CHIR with Wnt3a induced significant spontaneous differenti-ation and could not generate a homogenous NSC population,suggesting that GSK3 inhibition may coordinate multiple signalsbesides canonical Wnt activation in the context of neural in-duction under this condition. One possible explanation for thisspecific neural induction is that inhibition of TGF-β/Nodal sig-naling by SB431542 not only blocks the formation of mesen-doderm, but also engages in cross-talk with GSK3-mediatedsignaling (for example FGF signaling) to enhance neural in-duction, possibly by modulating a downstream component ofendogenous BMP signaling (2, 31). In addition, very recentstudies showed that GSK3 is a master regulator of in vivo neuralprogenitor homeostasis (28, 29). It is possible that neural in-duction is also coupled with the capture/maintenance of primitiveNSCs through GSK3 inhibition. Specifically, the combination ofGSK3 inhibitor, TGF-β receptor inhibitor, and hLIF is uniquelyrequired for long-term self-renewal of pNSCs under chemicallydefined conditions. Recent in vivo studies demonstrated thatTGF-β pathway activation counteracts canonical Wnt and nega-tively regulates self-renewal of midbrain neuroepithelial stemcells in the developing mouse brain (32). Loss of TGF-β signalingresults in neuroepithelial expansion in the midbrain, but not the

forebrain (32). The use of GSK3 inhibitor (which can activatecanonical Wnt) and TGF-β receptor inhibitor may partly re-capitulate such in vivo self-renewal signals of midbrain NSCs.With an improved understanding of the signaling mechanismsinvolved in lineage specification and maintenance of tissue-specificstem cells, this strategy could also be generalized and applied tothe capture of self-renewing stem cells from other germ layers,such as endoderm or mesoderm. Finally, this protocol also pro-vides a valuable tool with which to study the early molecularevents initiating human neural induction.

Materials and MethodsFor further details of cell cultures, neuronal differentiation, immunocyto-chemistry, flow cytometry, quantitative and semiquantitative RT-PCR, elec-trophysiological analysis, microarray analysis, teratoma assays, in vivotransplantation, and histology, see SI Materials and Methods. The antibodiesused in this study are shown in Table S1. For the primers of quantitative andsemi-quantitative RT-PCR, see Table S2.

ACKNOWLEDGMENTS. We thank our colleague Jem Efe for reading themanuscript and for providingmany insightful comments and suggestions, andJianwei Che (Genomics Institute of the Novartis Research Foundation, SanDiego) for analyzing the microarray data. S.D. and K.Z. are supported byNational Institutes ofHealth (NIH)Director’s Transformative R01 Program (R01EY021374). S.D. is supported by funding from the National Institute of ChildHealth and Development, the National Heart, Lung, and Blood Institute, andthe National Institute of Mental Health/NIH, the California Institute for Re-generative Medicine (CIRM), the Prostate Cancer Foundation, Fate Therapeu-tics, the Esther B. O’Keeffe Foundation, andThe Scripps Research Institute. K.Z.is supported by grants from the National Eye Institute/NIH, a Veteran AffairsMerit Award, the Macula Vision Research Foundation, Research to PreventBlindness, a Burroughs Wellcome Fund Clinical Scientist Award in Transla-tional Research, the Dick and Carol Hertzberg Fund, and Chinese National985 Project to Sichuan University andWest China Hospital. S.A.L. is supportedby grants from the National Eye Institute, the National Institute of Neurolog-ical Disorders and Stroke, the National Institute of Child Health and Develop-ment, the National Institute of Environmental Health Sciences, and CIRM.

1. Wobus AM, Boheler KR (2005) Embryonic stem cells: prospects for developmentalbiology and cell therapy. Physiol Rev 85:635–678.

2. Chambers SM, et al. (2009) Highly efficient neural conversion of human ES and iPScells by dual inhibition of SMAD signaling. Nat Biotechnol 27:275–280.

3. Elkabetz Y, Studer L (2008) Human ESC-derived neural rosettes and neural stem cellprogression. Cold Spring Harb Symp Quant Biol 73:377–387.

4. Elkabetz Y, et al. (2008) Human ES cell-derived neural rosettes reveal a functionallydistinct early neural stem cell stage. Genes Dev 22:152–165.

5. Pankratz MT, et al. (2007) Directed neural differentiation of human embryonic stemcells via an obligated primitive anterior stage. Stem Cells 25:1511–1520.

6. Tropepe V, et al. (2001) Direct neural fate specification from embryonic stem cells:a primitive mammalian neural stem cell stage acquired through a default mechanism.Neuron 30:65–78.

7. Conti L, Cattaneo E (2010) Neural stem cell systems: physiological players or in vitroentities? Nat Rev Neurosci 11:176–187.

8. Qian X, et al. (2000) Timing of CNS cell generation: a programmed sequence of neuronand glial cell production from isolated murine cortical stem cells. Neuron 28:69–80.

9. Koch P, Opitz T, Steinbeck JA, Ladewig J, Brüstle O (2009) A rosette-type, self-renewing human ES cell-derived neural stem cell with potential for in vitro instructionand synaptic integration. Proc Natl Acad Sci USA 106:3225–3230.

10. Ring DB, et al. (2003) Selective glycogen synthase kinase 3 inhibitors potentiateinsulin activation of glucose transport and utilization in vitro and in vivo. Diabetes 52:588–595.

11. Sato N, Meijer L, Skaltsounis L, Greengard P, Brivanlou AH (2004) Maintenance ofpluripotency in human and mouse embryonic stem cells through activation of Wntsignaling by a pharmacological GSK-3-specific inhibitor. Nat Med 10:55–63.

12. Li W, et al. (2009) Generation of rat and human induced pluripotent stem cells bycombining genetic reprogramming and chemical inhibitors. Cell Stem Cell 4:16–19.

13. Tojo M, et al. (2005) The ALK-5 inhibitor A-83-01 inhibits Smad signaling and epithelial-to-mesenchymal transition by transforming growth factor-beta. Cancer Sci 96:791–800.

14. Smukler SR, Runciman SB, Xu S, van der Kooy D (2006) Embryonic stem cells assumea primitive neural stem cell fate in the absence of extrinsic influences. J Cell Biol 172:79–90.

15. Hitoshi S, et al. (2004) Primitive neural stem cells from the mammalian epiblastdifferentiate to definitive neural stem cells under the control of Notch signaling.Genes Dev 18:1806–1811.

16. Hitoshi S, et al. (2002) Notch pathway molecules are essential for the maintenance,but not the generation, of mammalian neural stem cells. Genes Dev 16:846–858.

17. Seiffert D, et al. (2000) Presenilin-1 and -2 are molecular targets for γ-secretaseinhibitors. J Biol Chem 275:34086–34091.

18. Yao S, et al. (2006) Long-term self-renewal and directed differentiation of humanembryonic stem cells in chemically defined conditions. Proc Natl Acad Sci USA 103:6907–6912.

19. Ellis P, et al. (2004) SOX2, a persistent marker for multipotential neural stem cellsderived from embryonic stem cells, the embryo or the adult. Dev Neurosci 26:148–165.

20. Marzesco AM, et al. (2005) Release of extracellular membrane particles carrying thestem cell marker prominin-1 (CD133) from neural progenitors and other epithelialcells. J Cell Sci 118:2849–2858.

21. Corbeil D, et al. (2000) The human AC133 hematopoietic stem cell antigen is alsoexpressed in epithelial cells and targeted to plasma membrane protrusions. J BiolChem 275:5512–5520.

22. Gammill LS, Bronner-Fraser M (2003) Neural crest specification: migrating intogenomics. Nat Rev Neurosci 4:795–805.

23. Zhang SC (2006) Neural subtype specification from embryonic stem cells. Brain Pathol16:132–142.

24. Seri B, et al. (2006) Composition and organization of the SCZ: a large germinal layercontaining neural stem cells in the adult mammalian brain. Cereb Cortex 16(Suppl 1):i103–i111.

25. Smidt MP, et al. (1997) A homeodomain gene Ptx3 has highly restricted brainexpression in mesencephalic dopaminergic neurons. Proc Natl Acad Sci USA 94:13305–13310.

26. Joksimovic M, et al. (2009) Spatiotemporally separable Shh domains in the midbraindefine distinct dopaminergic progenitor pools. Proc Natl Acad Sci USA 106:19185–19190.

27. Tang M, et al. (2010) Interactions of Wnt/beta-catenin signaling and sonic hedgehogregulatetheneurogenesisofventralmidbraindopamineneurons. JNeurosci30:9280–9291.

28. Mao Y, et al. (2009) Disrupted in schizophrenia 1 regulates neuronal progenitorproliferation via modulation of GSK3beta/beta-catenin signaling. Cell 136:1017–1031.

29. Kim WY, et al. (2009) GSK-3 is a master regulator of neural progenitor homeostasis.Nat Neurosci 12:1390–1397.

30. Hur EM, Zhou FQ (2010) GSK3 signalling in neural development. Nat Rev Neurosci 11:539–551.

31. Gerrard L, Rodgers L, Cui W (2005) Differentiation of human embryonic stem cells toneural lineages in adherent culture by blocking bone morphogenetic proteinsignaling. Stem Cells 23:1234–1241.

32. Falk S, et al. (2008) Brain area-specific effect of TGF-beta signaling on Wnt-dependentneural stem cell expansion. Cell Stem Cell 2:472–483.

8304 | www.pnas.org/cgi/doi/10.1073/pnas.1014041108 Li et al.

Dow

nloa

ded

by g

uest

on

June

15,

202

1

http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=STXThttp://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=ST1http://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1014041108/-/DCSupplemental/pnas.201014041SI.pdf?targetid=nameddest=ST2www.pnas.org/cgi/doi/10.1073/pnas.1014041108