EMBRYONIC STEM CELLS/INDUCED PLURIPOTENT STEM CELLS

Long-Term, Stable Differentiation of Human Embryonic StemCell-Derived Neural Precursors Grafted into the AdultMammalian Neostriatum

IGOR NASONKIN,a VASILIKI MAHAIRAKI,a LEYAN XU,a GLEN HATFIELD,a BRIAN J. CUMMINGS,d,e CHARLES EBERHART,f

DAVID K. RYUGO,g DRAGAN MARIC,h ELI BAR,a VASSILIS E. KOLIATSOSa,b,c

aDepartment of Pathology, Division of Neuropathology, bDepartment of Neurology, and cDepartment of Psychiatryand Behavioral Sciences, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA;dPhysical Medicine & Rehabilitation and eReeve-Irvine Research Center, University of California Irvine, IrvineCalifornia, USA; fDepartments of Ophthalmology and Oncology and gDepartments of Otolaryngology andNeuroscience, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA; hFlow CytometryCore Facility, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda,Maryland, USA

Key Words. Cellular therapy • Embryonic stem cells • Neural differentiation • Neural induction • Neural stem cells • Pluripotent stem cells• Stem cell plasticity • Stem cell transplantation

ABSTRACT

Stem cell grafts have been advocated as experimentaltreatments for neurological diseases by virtue of their abil-ity to offer trophic support for injured neurons and, theo-retically, to replace dead neurons. Human embryonic stemcells (HESCs) are a rich source of neural precursors (NPs)for grafting, but have been questioned for their tendencyto form tumors. Here we studied the ability of HESC-derived NP grafts optimized for cell number and differen-tiation stage prior to transplantation, to survive and stablydifferentiate and integrate in the basal forebrain (neostria-tum) of young adult nude rats over long periods of time(6 months). NPs were derived from adherent monolayercultures of HESCs exposed to noggin. After transplanta-tion, NPs showed a drastic reduction in mitotic activityand an avid differentiation into neurons that projected via

major white matter tracts to a variety of forebrain targets.A third of NP-derived neurons expressed the basal fore-brain-neostriatal marker dopamine-regulated and cyclicAMP-regulated phosphoprotein. Graft-derived neuronsformed mature synapses with host postsynaptic structures,including dendrite shafts and spines. NPs inoculated inwhite matter tracts showed a tendency toward glial (pri-marily astrocytic) differentiation, whereas NPs inoculatedin the ventricular epithelium persisted as nestin(1) pre-cursors. Our findings demonstrate the long-term ability ofnoggin-derived human NPs to structurally integrate tu-mor-free into the mature mammalian forebrain, whilemaintaining some cell fate plasticity that is strongly influ-enced by particular central nervous system (CNS) niches.STEM CELLS 2009;27:2414–2426

Disclosure of potential conflicts of interest is found at the end of this article.

INTRODUCTION

Neural stem cells presently considered as cell therapies forneurological disorders are derived primarily from adult or fe-tal neural tissue or from in vitro differentiated embryonicstem cells [1–3]. Although both methods of derivation haveadvantages and disadvantages and new methodologies involv-ing somatic cell derivation are being developed [4], thehuman embryonic stem cell (HESC) approach continues toprovide a theoretically inexhaustible and very pliable source

of cells [5]. The plasticity and differentiation potential ofHESCs have been demonstrated in recent in vitro studieswhere these cells were induced to evolve into specific neuro-nal and glial lineages to become mesencephalic dopaminergicneurons [6–8], motor neurons [9–11], and oligodendrocytes[12]. The traditional ES cell differentiation paradigm involvesthe formation of embryoid bodies [13, 14] where cells evolvein a fashion similar to that in the embryonic gastrula; that is,they go through the three main embryonic germ layer stages.These cultures contain high admixtures of non-neuroepithelialcells [15], a condition that may be a disadvantage for neural

Author contributions: I.N.: design and performance of most experiments, manuscript writing; V.M., L.X., G.H., B.J.C., E.B., D.M., andD.R.: performance of some experiments, manuscript writing; C.E.: design of some experiments, manuscript writing; V.E.K.: conceptionand design, financial support, manuscript writing and editing.

Correspondence: Igor Nasonkin, Ph.D., NIH/NEI, Neurobiology, Neurodegeneration & Repair Laboratory (N-NRL), 9000 RockvillePike, Bldg. 6, Rm 341, Bethesda, MD 20892-0610, USA. Telephone: 617-388-4104; Fax: 301-480-1769; e-mail: nasonkini@mail.nih.gov; or Vassilis E. Koliatsos, M.D., The Johns Hopkins University School of Medicine, Division of Neuropathology, 558 Ross ResearchBuilding, 720 Rutland Avenue, Baltimore, Maryland 21205, USA. Telephone: 410-502-5172; Fax: 410-955-9777; e-mail: koliat@jhmi.edu Received January 28, 2009; accepted for publication June 24, 2009; first published online in STEM CELLS EXPRESS July 16, 2009.VC AlphaMed Press 1066-5099/2009/$30.00/0 doi: 10.1002/stem.177

STEM CELLS 2009;27:2414–2426 www.StemCells.com

replacement therapies that require more defined cell popula-tions for transplantation. A new generation of protocolsdesigned to avoid embryoid body formation has achieved agreater degree of in vitro commitment of HESCs to neuroepi-thelial fates [16, 17]. Further progress was made with theincorporation of protocols designed to mimic neural inductionevents that occur early in development [18], including thoseutilizing adherent monolayer conditions in which bone mor-phogenetic protein (BMP) signaling is blocked by noggin [19,20]. These cultures allow for the homogeneous exposure ofHESCs to the neural lineage-promoting action of noggin andcan reliably generate neural precursors (NPs) for furtherdifferentiation.

To test the generic potential of noggin-differentiatedHESCs for therapeutic applications in the nervous system, weevaluated the in vivo outcomes of grafting these cells in theadult mammalian forebrain. In the present study, we demon-strate the consistent, controlled differentiation of HESCs toNPs using noggin without additional inducing agents ortrophic factors and the ability of such derived NPs to inte-grate, to various degrees, into the adult mammalian forebrain.Within the course of several months explored in this study,HESC-derived NPs undergo substantial neuronal differentia-tion without apparent tumorigenesis, project axons to distantlocations, and form structurally mature synapses in theirinnervation targets.

METHODS

Propagation and Neuroepithelial Differentiationof HESCsThe HESC line BG01, included in the NIH Human EmbryonicStem Cell Registry (http://stemcells.nih.gov/research/registry),was obtained from BresaGen, Inc., Athens, GA (http://www.novocell.com/) [5]. HESC colonies were grown on mitoticallyinactivated mouse embryonic fibroblasts (MEFs; 2 ! 106 cellsper 35-mm dish) prepared from E12 ICR mice (Taconic, Hudson,NY, http://www.taconic.com). Culturing conditions were accord-ing to company guidelines, except that higher concentrationsof bFGF (10 ng/ml, Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) and 15% knockout replacement serum (Invi-trogen, Carlsbad, CA, http://www.invitrogen.com) were used.Colonies were passaged every 4–7 days by manual dissection.For neural induction, days 4–6 HESC colonies growing on MEFswere treated with a medium composed of 50% fresh HESC me-dium (see above) and 50% Neurobasal (NB) medium (Invitro-gen), supplemented with 100 ng/ml noggin (Sigma-Aldrich).Over the next 2 days, noggin concentration was graduallyincreased (day 2, 200 ng/ml; day 3, 250 ng/ml) and, on day 3,colonies were mechanically dislodged from the MEF layer andtransferred to gelatin-coated tissue culture plates in NB completemedium. At 4 weeks, cultures were transferred to new gelatin-coated dishes (passage 1, P1) in NB complete medium withbFGF and noggin. NPs were passaged once again (P2) and thenused for transplantation or were frozen. Details are provided inthe supporting information data.

Monitoring Differentiation of HESCs In VitroQuantitative Real-Time Polymerase Chain Reaction.Undifferentiated HESC colonies (days 4–5 after passaging) grow-ing on MEFs were collected by mechanical dissection and placedinto Buffer RLT (Qiagen, Valencia, CA, http://www1.qiagen.-com). Total RNA was prepared according to manufacturer’sinstructions. HESC colonies differentiating on gelatin-coated 35-mm dishes were lysed directly on the plates on days 7, 14, 21,

and 28 and at P2 (1–2 plates per time point) using Buffer RLT.Total RNA was extracted from the lysates using an RNeasy kit(Qiagen) and was treated with DNase according to manufacturer’sprotocol. Real-time polymerase chain reaction (RT-PCR) was per-formed essentially as described [2] (supporting information data).

Immunocytochemistry. NP cultures were rinsed with phos-phate-buffered saline and briefly fixed with 4% freshly depoly-merized, neutral-buffered paraformaldehyde for 15 minutes. Cul-tures were processed for the immunocytochemical detection ofneural stem cell and differentiation protein markers as per detailsdescribed in supporting information data. Both single and multi-ple epitope immunocytochemistry (ICC) were performed andcombined with DAPI counterstain. Total number of cells in 10randomly selected fields was counted with the help of an auto-mated cell-counting program (AxioVision; Carl Zeiss, Jena, Ger-many, http://www.zeiss.com).

Immunophenotyping of NPs with Flow Cytometry. Cellswere first dispersed into a uniform single-cell suspension usingpapain digestion protocol as described elsewhere [21]. The result-ing mixture of cells was immunolabeled with the following cock-tail of lineage-selective surface markers: rabbit IgG1 anti-humanCD133 and mouse IgM anti-human CD15 (Santa Cruz Biotech-nology Inc., Santa Cruz, CA, http://www.scbt.com), mouse IgG1anti-human CD29 (BD Biosciences, San Jose, CA, http://www.bdbiosciences.com), and a mixture of tetanus toxin frag-ment C (TnTx)-mouse anti-TnTx IgG2b. Methodological detailsare included in supporting information data.

Animals and Surgical ProceduresNude rats (strain CR: NIH-rnu; male; 6–8 weeks old; n " 22)were purchased from the National Cancer Institute. All surgicalprocedures conformed to protocols approved by the Animal Careand Use Committee of the Johns Hopkins Medical Institutionsand were carried out by using gas anesthesia (enflurane:oxygen:-nitrous oxide " 1:33:66) and aseptic methods. NPs were injectedin suspension (15 ! 103 cells in 1 ll) into the neostriatum of ani-mals mounted on a Kopf stereotaxic unit (David Kopf Instru-ments, Tujunga, CA, http://www.kopfinstruments.com). Injectionswere made under microscopic guidance via pulled glass micro-pipettes controlled by a Nanoinjector device (World PrecisionInstruments, Sarasota, FL, http://www.wpiinc.com). Animals werekilled with perfusion-fixation as described elsewhere [2] at 1.5months (n " 8), 3 months (n " 7), and 6 months (n " 7).

The number of cells grafted was based on a pilot experimentdesigned to optimize integration of graft with host tissues and tomonitor tumorigenesis as a function of graft size (see supportinginformation data, pilot experiment, and supporting informationFig. 1A–1E). In this pilot, we varied the number of cells(between 20 ! 103 and 120 ! 103) grafted in a single inoculationsite into the neostriatum. Variance in the number of injected cellsbetween 20 ! 103 and 120 ! 103 appeared to have little effecton the degree of tissue integration or the mitotic activity ofgrafts. Human-specific synaptophysin immunoreactivity was seenin small patches of grafted cells at 1.5 months, but increased sub-stantially by 3 months and was seen throughout the graft at 6months postgrafting (supporting information Fig. 2). Although theresults of these pilot studies were reassuring with respect to gen-eral histopathology and tumorigenesis, the transient appearance ofneuroepithelial phenotypes led us to decrease the number of cellsin the graft to 15 ! 103.

Monitoring Differentiation of HESCs In Vivo:Histology and ICCThe survival, growth, and phenotypic fate of HESC-derived NPsin vivo were assessed with immunoperoxidase or dual-label im-munofluorescence. Tissues were prepared from animals perfused

Nasonkin, Mahairaki, Xu et al. 2415

www.StemCells.com

with 4% freshly depolymerized paraformaldehyde as describedelsewhere [2]. Brain blocks were sectioned (30 lm) in the coro-nal or sagittal planes. Primary antibodies were used to disclosehuman (graft) versus rat (host) cell identity, mitotic activity, andneuronal and glial phenotype specification and included a numberof well-characterized monoclonal antibodies and antisera (sup-porting information Table 1). Normal IgG from the species of ori-gin of primary antibodies served as negative controls.

For immunoperoxidase staining, sections were processed withperoxidase-antiperoxidase or avidin-biotin-peroxidase [22]. Dual-label immunofluorescence was used to study NP mitotic activity

and differentiation, combined immunoreactivity for human nu-clear antigen (HNu) and that for a cell proliferation or differentia-tion marker, and was performed essentially as described else-where [2]. ICC controls were prepared by replacing primaryantibodies with pre-immune serum from the same species of ori-gin and were negative in all cases. To measure rates of differen-tiation of HESC-derived NPs, we used two representative sectionsthrough the middle of each graft and counted the total number ofHNu(#) cells, and also the cells dually labeled with HNu andone of the chosen phenotypic markers on 10 randomly selected!100 fields per case. The total number of HNu(#) cells and

Figure 1. Neural induction/differentiation ofHESCs with noggin in adherent monolayer, fromday 0 to P2. (A): Appearance of undifferentiatedHESCs on MEFs at day 0 by phase contrast. (B):Induction/differentiation protocol used to generateNPs used in this study. (C): Day 0 (undifferenti-ated) HESCs express, by ICC, index pluripotencymarkers including Oct3/4, SSEA4, Tra1-60, andTra1-81. (D): Oct3/4 immunoreactivity is substan-tially reduced in differentiating HESCs betweenday 9 (left) and day 28 (right) in culture, evidenceof shedding of HESC pluripotency. Scale bars:20 lm.

2416 Differentiation of HESC-Derived Neural Precursors

double-labeled profiles were pooled for each case, and percen-tages of double-labeled cells were generated per case and thengrouped per time point designation (1.5, 3, or 6 months). Varia-tion in differentiation, as a function of time point after surgery,was studied with one-way ANOVA followed by Tukey’s multiplecomparison post hoc test. Results are expressed as mean $ stand-ard error.

Electron MicroscopyFollowing transcardiac perfusion with 4% freshly depolymerizedparaformaldehyde, brains were removed and embedded in a gela-tin-albumin mixture hardened with glutaraldehyde. Vibratomesections were subjected to synaptophysin ICC, embedded, andresectioned at the ultrathin level (supporting information data).

RESULTS

Controlled Differentiation of HESCs to NPs In VitroHESC colonies were grown on a MEF layer supplementedwith bFGF and differentiated to neuroepithelial fate with nog-gin (Fig. 1A, 1B). The pluripotency markers Oct3/4, SSEA4,Tra1-60, and Tra1-81 were amply expressed in these startingHESC colonies (Fig. 1C, 1D). After HESC colonies were dis-lodged from the MEF layer, they attached almost immediatelyto gelatinized plates and showed robust growth and flatappearance. Ten days after transfer, multiple early rosettesappeared at the center of the colonies. The cultures acquiredthree-dimensional growth at 15 days after transfer; at thistime point, HESC colonies grew in the form of early rosettes(i.e., three-dimensional structures comprised of columnar epi-thelial cells) and proliferated vigorously. In 3 weeks, early ro-sette structures became less well defined and, by 1 month,most of them disappeared and cultures acquired two-dimen-sional neuroepithelial characteristics.

Differentiation of HESCs to NPs was monitored with RT-PCR, ICC, and flow cytometry (Figs. 1–3). RT-PCR datademonstrate that, with ongoing noggin treatment, there is agradual upregulation of neural markers and a downregulationof pluripotency markers such as Oct3/4 (Figs. 1D, 2A), Nanog(Fig. 2A), and SSEA4 (data not shown). The stem cell markerSox2 continues to be expressed in differentiating HESCs.Neural markers show gradual upregulation, with most of themrising between day 7 and day 21; further upregulation of neu-ral markers is detected at passage 2 after the replating of NPs(Fig. 2A). The endodermal marker GATA-4 becomes unde-tectable by passage 2 (Fig. 2A). The neural crest markerMsx1 shows gradual upregulation (Fig. 2A), whereas twoadditional markers of mesoderm and endoderm (Brachyuryand Pdx1, respectively) are absent during the differentiationprocess (data not shown). The expression of a panel of neuro-trophic factors including the neurotrophins NGF and BDNF,GDNF, VEGF, and bFGF increases between day 14 and day21 and then decreases or reaches a plateau (Fig. 2B). A repre-sentative panel of transcription factors conveying anterior,posterior, and dorsal-ventral position specification in the neu-ral tube failed to demonstrate a predominant pattern in cul-tured NPs, despite some up- or downregulation of individualtranscripts over time (Fig. 2C).

HESC-derived NPs were profiled by ICC immediatelyprior to transplantation (Fig. 3A–3E). More than 80% of thecells was immunoreactive for nestin (Fig. 3A). Seven percentof the cells expressed the class III b-tubulin epitope TUJ1,serving as a marker of immature neurons (Fig. 3B). Six per-cent of the cells expressed the astrocytic marker GFAP (Fig.3C), whereas 10% of the cells expressed S100b, a marker of

ependymal, radial glial, and early astrocytic cells (Fig. 3D).Multiple-antigen ICC confirmed further the predominant phe-notypes of P2 NPs (Fig. 3E). For example, nestin was foundto colocalize extensively with musashi-1 and vimentin, a pat-tern confirming the neural stemness of our NPs, whereas thesecells were negative for pax6, an early marker of neuroepithe-lium. The neural stem cell marker CD15 (SSEA-1) wasexpressed by approximately 15% of NPs, whereas the com-mon glial precursor marker A2B5 was present in less than 1%of NPs. Approximately 7% of the cells was mitotically activeby Ki67 immunostaining (data not shown).

Immunophenotyping of P2 NPs by flow cytometry showedthat less than 1% of the total NPs express the surface stemcell marker CD133 that denotes the presence of symmetricaldivision (Fig. 3F). In contrast, the great majority (approxi-mately 79%) of NPs were positive for tetanus toxin (TnTx), amarker of neuronal progenitors [23, 24]. CD15#TnTx% cellscomprised 16.80% of our NP population. A small subset(4.3%) of NPs was CD15%TnTx% (Fig. 3F-2) and 14.5% ofthese CD15%TnTx% NPs was CD29# cells (Fig. 3F-3).

The above findings demonstrate the consistent derivationof multipotential NPs from HESCs by gradual differentiation.A combination of methodologies including ICC and immuno-phenotyping indicate that, at the time of grafting into brain,the majority of our NPs were at the stage of very early neuro-nal progenitors [nestin(#) and TnTx(#) cells], whereas a size-able one fifth or so of our NP population was at an earlierneural stem stage (CD15#TnTx% cells), capable of generatingneurons and glial cells.

Survival and Proliferation of HESC-DerivedNPs in the NeostriatumTo investigate the in vivo potential of HESC-derived NPs tointegrate into the adult mammalian brain, we grafted P2 NPsinto the neostriatum of young adult nude rats. HNu-immuno-stained material demonstrated the successful engraftment andexcellent survival of NPs in this forebrain site (Fig. 4A–4C),with few scattered apoptotic profiles detected only in 1.5-month grafts. In some 1.5-month grafts, clusters of packedbasophilic cells were present, but these were not observed atlater time points. Because of the well-known propensity ofHESC-derived grafts for tumorigenesis [25], we tested graftsfor the expression of the nuclear proliferation marker Ki67 at1.5, 3, and 6 months after transplantation (Fig. 4D, 4E). TheKi67 proliferation index within the grafts decreased from8.3% $ 2.4% at 1.5 months (n " 8) to 4.2% $ 0.7% at3 months (n " 7) and then to 1.2% $ 0.6% at 6 months (n "7), suggesting a gradual exit from the cell cycle and differen-tiation of grafted NPs.

Consistent with a pattern of progressive maturation overtime, the distinction of grafts from surrounding normal braintissue, although pronounced at 1.5 months, gradually disap-peared as the grafts assumed a more homogeneous appearance(supporting information Fig. 1). No teratomas [26], neuraltumors [27], or clusters of Ki67 immunoreactive cells wereobserved within the grafts, findings that indicate the absenceof tumorigenic proliferative masses.

Plasticity, Migration, and Neuronal Differentiationof HESC-Derived NPsIn the primary grafting location (neostriatum), the majority ofNPs differentiated into TUJ-1-positive neurons (Fig. 5). Closeto 50% (44.6% $ 3.3%) of HNu(#) cells were nestin-positiveat 1.5 months, a rate reduced to approximately 10% (8.5% $1.4%) at 3 months and less than 1% (0.2% $ 0.1%) at6 months after grafting. A large number of HNu(#) cells

Nasonkin, Mahairaki, Xu et al. 2417

www.StemCells.com

Figure 2. Monitoring differentiation of HESCs in vitro by RT-PCR. Relative gene expression is shown as fold change in gene expression usingthe comparative Ct (DDCt) method. Expression levels of genes in the initial HESC sample were used as standard (equal to 1) to which all otherexpression values were compared and expressed as fold changes. Bars corresponding to data from P2 NPs are shown in darker colors, to distin-guish them from data derived from day 7 to day 28 HESC colonies that continuously differentiate within the same plates. (A): Expression of indexpluripotency and neural markers in differentiating HESCs, measured by quantitative RT-PCR at days 7, 14, 21, 28, and 49 (P2). Stem cell markersare shown in yellow, early neural markers are in green, neuronal and glial markers are in gray, neural crest markers are in blue, and endodermalmarkers are in pink. The expression of mesodermal marker Brachyury was not detectable and, therefore, is not shown. (B): Expression of trophicfactors including neurotrophins (NGF, BDNF, GDNF, VEGF, and bFGF) by differentiating HESCs at days 7, 14, 21, 28, and 49. (C): Expressionof selected dorsal-ventral (blue), anterior (pink), and posterior (purple) markers by differentiating NPs at day 28 and P2. Several markers (Otx2,Emx2, Dlx2, and Nkx2.1) were not detectable and therefore are not shown. Abbreviations: NGF, nerve growth factor; BDNF, brain-derived neuro-trophic factor; GDNF, glial-derived neurotrophic factor; VEGF, vascular endothelial growth factor; bFGF, basic fibroblast growth factor.

2418 Differentiation of HESC-Derived Neural Precursors

within the graft expressed doublecortin, a marker for migrat-ing neuroblasts and early neurons [28, 29] (Fig. 5A). Morethan one half of HNu(#) cells were doublecortin(#) at 1.5months (52.1% $ 3.4%) and 3 months (69.6% $ 5.3%); asimilar percentage of HNu(#) cells also expressed the double-cortin-like kinase epitope DCAMKL-1 [72.0% $ 3.4% of

HNu(#) cells] at 3 months after grafting. Doublecortinexpression was reduced to less than 5% (4.7% $ 0.7%) ofHNu(#) cells at 6 months after grafting (Fig. 5C). On the ba-sis of a sharp rise in TUJ-1 colocalization of HNu(#) between1.5 months (20.1% $ 3.9%) and 3 months (70.8% $ 6.8%)after grafting (Fig. 5B, 5C), we assume that neuronal fate was

Figure 3. Immunocytochemical (ICC) andimmunophenotypic profiling of HESC-derivedNPs at the time of transplantation. (A–D): Thesesingle-epitope ICC preparations illustrate theexpression of selective neural stem, neuronal,and glial markers. Insets represent magnifica-tions of areas shown with asterisks in themerged image. Although the vast majority ofcells express the neural stem cell marker nestin(A), one tenth or less of them express the classIII b-tubulin epitope TUJ1, a marker of imma-ture neurons (B), or the glial (primarily astro-glial) markers GFAP (C) and S100b (D). (E):These four-epitope ICC preparations serve tofurther characterize the phenotype of HESC-derived NPs at the time of transplantation. Epi-tope combinations are nestin-vimentin-A2B5-pax6 (top) and nestin-vimentin-musashi1-CD15(bottom). Nestin colocalized extensively withvimentin (top, bottom) and musashi (bottom).Pax6, revealed with Alexa Fluor 750, was notpresent in these cells and is not shown here. Asubpopulation of NPs expressed the early neuro-nal precursor marker CD15 (bottom), whereasvery few expressed A2B5 (top). Asterisks indi-cate areas further enlarged in the insets. (F):Immunophenotyping of HESC-NPs by flowcytometric analysis. HESC-derived NPs weredispersed into a uniform single-cell suspensionand labeled with antibodies targeting CD133,CD29, CD15, and TnTx. Dead cells were coun-terstained with DAPI and the resulting vital phe-notypes assayed by flow cytometry. Data arepresented as two-parameter dot density plots inpseudocolor that reveal the relative abundance(percentiles) of different cell populations byusing either electronically gated regions as in(1) (Live Cells, Dead Cells, Debris) or quadrantsas in (2, 3). In (1), DAPI was used in conjunc-tion with forward scatter to discriminatebetween live (DAPI-negative) and dead (DAPI-positive) cells. Data in (2, 3) are based on livecells, which represent approximately 80% of thepreparation. In (2), the great majority of livecells are TnTx# (approximately 79%, that is, thesum of 48.21 and 30.73). CD15 labeling furthersubdivides these neuronal-restricted precursorsinto CD15# (30.73%) and CD15% (48.21%)subpopulations. CD15#TnTx% NPs comprise16.80% of total NPs. The CD133 versus CD29plot in (3) is gated on 4.26% of CD15%TnTx%

cells (see panel F 2). Only 14.47% of these cellsare CD29#, which amounts to approximately1% of total live cells (14.47 ! 4.26/100).CD133# cells represent less than 1% (0.75%) ofthe total population of live cells sorted. Abbrevi-ations: Scale bars: 50 lm.

Nasonkin, Mahairaki, Xu et al. 2419

www.StemCells.com

Figure 4. Transplantation strategy and survival of human NPs in rat neostriatum. (A): This is a composite sagittal map of individual striatalgrafts at maximal inoculation size. Contours of each individual graft are shown in black. Contours of rat brain and the ventricular lining are inbold gray. The map shows a reliable coverage of a large section of anterior caudate-putamen with rare invasion into the adjacent globus pallidus.Major anatomical landmarks are depicted for orientation purposes (see abbreviations below). (B): This coronal section through the anterior sep-tum shows a typical striatal graft of human NPs 3 months after inoculation. (C): This is an HNu-stained sagittal section through the middle ofHESC-derived NP graft 3 months after grafting. The section was counterstained with cresyl violet. Most HNu(#) cells have large round nucleiconsistent with mature neurons. The inset represents the magnification of framed area. (D, E): Mitotic activity within HESC-derived NP grafts asshown by dual staining for HNu and Ki67. (D): Mitotic profiles of NPs in grafts measured at 1.5, 3, and 6 months. Bars indicate group averages$ SEM. Variance in mitotic profiles between the three cohorts was significant (ANOVA, p " .014). Significant post hoc differences are indicatedwith brackets. (E): A representative area within a graft 3 months after transplantation, dually stained with HNu (red nuclear marker) and Ki67(green nuclear marker). The three images illustrate the low rate of mitotic activity in HESC-derived grafts. The single double-stained nuclear pro-file (bottom) is indicated with an asterisk. Abbreviations: Acb, nucleus accumbens; AOD, anterior olfactory nucleus, dorsal; AOV, anterior olfac-tory nucleus, ventral; Cx, cortex; cc, corpus callosum; GP, globus pallidus; Hy, hypothalamus; STh, subthalamic nucleus; Str, corpus striatum(neostriatum); th, thalamus; VP, ventral pallidum. *, p < .05. Scale bars: (B, C): 100 lm (main panel), 10 lm (inset); (E): 10 lm.

2420 Differentiation of HESC-Derived Neural Precursors

established, for the majority of grafted NPs, prior to 3 monthsafter grafting. Further increase of TUJ-1 expression between 3and 6 months (to 86.6% $ 1.8%) was not significant.

Human-specific synaptophysin immunostaining was seenin islands of grafted cells at 1.5 months, but increased substan-tially by 3 months and was seen throughout the graft at6 months after grafting (supporting information Fig. 2). Therewas no GAD or ChAT colocalization with HNu(#) profiles orhuman-specific synaptophysin(#) terminals in 6-month grafts.Only sparse colocalization of GluR2/3 and HNu was seen ingrafts and there was no visible VGLUT1 or VGLUT2 immu-noreactivity in human-specific synaptophysin(#) terminals. Anumber of markers of basal forebrain and basal ganglia neu-rons were expressed by differentiating NPs. DARPP-32 wasnot detected in 3-month grafts, but was expressed by nearlyone third (30.1% $ 3.2%) of HNu(#) cells at 6 months aftergrafting (supporting information Fig. 3A–3C). The calcium-binding protein calretinin was expressed by less than 1% ofHNu(#) cells at 3 months, but was present in 6.5% $ 1.6% ofHNu(#) cells at 6 months after grafting (supporting informa-tion Fig. 3D, 3D0). Parvalbumin, another calcium-binding pro-tein, was undetectable in 1.5- or 3-month grafts and was foundin less than 1% of HNu(#) cells at 6 months after grafting.

In a few cases in which grafting inadvertently involvedlarge white matter tracts of the forebrain, that is, the fornix,the internal capsule, or the genu of the corpus callosum, amajority of NPs had astroglial phenotypes as judged by con-focal analysis of GFAP immunoreactivity in HNu(#) profilesand also with human-specific SC121 immunoreactivity(Fig. 6). In these cases, dual ICC for HNu and the oligodendro-cyte markers O4 or RIP also revealed small numbers (less than5%) of oligodendrocytes differentiating from grafted NPs.

Very few HNu(#) cells were found in parenchymal sitesaway from the graft. The only exception was the ventral fore-brain pia which, in virtually all cases examined, was denselypopulated by HNu(#) bipolar cells; all these cells expresseddoublecortin (supporting information Fig. 4), but were nega-tive for nestin and did not display neuronal configurations;that is, they had no discernible axons by SC121 ICC. How-ever, when HESC-derived NPs were spuriously depositedalong the needle track and became part of the host ventricularepithelium, they all maintained a nestin-positive state monthsafter the injection and none expressed doublecortin (support-ing information Fig. 5).

In concert, our dual ICC experiments with graft-specificand differentiation markers indicate a predominantly neuronaldifferentiation of HESC-derived NPs in the neostriatal graft-ing sites. When cells were inoculated into white matter tracts,fate choice was astrocytic and, in some cases, oligodendro-cytic. However, when human NPs were deposited in the epen-dymal region, they persisted in a neural stem cell state. Theprevious patterns demonstrate a fate-choice plasticity ofgrafted human NPs that may depend on host microenviron-ment. Parenchymal migration of those NPs was difficult to es-tablish; pial neuroblast profiles consistently observed in allgrafted cases could have arisen either from pial or via micro-vessel migration.

Evidence for Circuit Integration of HESC-DerivedNPs in the Rat Forebrain: Axons and SynapsesAxonal projections from the graft were visualized as early as1.5 months with doublecortin and DCAMKL-1 antibodies andat 3 and 6 months after grafting with human synaptophysinand human cytosol-specific SC121 antibodies (Fig. 7A, 7B).Human NF70(#) axons were evident only at 6 months aftergrafting (data not shown). These projections appeared as

Figure 5. Neuronal differentiation of HESC-derived NPs in rat neo-striatum. Photomicrographs represent cases of neuronal differentiationof HNu(#) cells by confocal microscopy from subjects prepared3 months after grafting. (A): Dual ICC for HNu and doublecortinshows many HNu(#) cells that are also immunoreactive for double-cortin (DCX). (B): Double ICC for HNu and TUJ1 demonstrates theabundance of double-stained cells in the grafts. (C): Bar graphsdepicting neuronal cell fate choice in striatal grafts of HESC-derivedNPs at 1.5, 3, and 6 months represented by percentages of DCX-la-beled (left panel) and TUJ1-labeled (right panel) HNu(#) cells. Barsindicate group averages $ SEM. Variance between the three cohortswas significant for both DCX (ANOVA, p " .0003) and TUJ1(ANOVA, p " .0004). Significant post hoc differences are indicatedwith brackets. *, p|<|.05; **, p|<|.001. Scale bars: 10 lm.

Nasonkin, Mahairaki, Xu et al. 2421

www.StemCells.com

dense bundles of axons traversing most white matter tracts inthe broad vicinity of the graft including the striatopallidalpencils, stria terminalis, internal capsule, anterior commissure(olfactory and temporal limbs), the forceps minor and body ofthe corpus callosum, and furthermore, the cerebral peduncleand callosal radiations.

Numerous human synaptophysin(#) boutons were foundin the surroundings of the graft as early as 3 months aftergrafting; some of these structures made contact with other dif-ferentiated NPs and these graft-derived cell-to-cell contactsbecame even denser at 6 months (supporting informationFigs. 2, 3). Synaptophysin(#) boutons away from the graftwere evident at 6 months after grafting in layer I of frontalcortex, the olfactory bulb (internal plexiform layer), anteriorolfactory nucleus, bed nucleus-central amygdala inserts in thesubstantia innominata, the globus pallidus, substantia nigra-pars reticulata, and the subthalamic nucleus (Fig. 7C).

Ultrastructural analysis of tissues from animals at 3 and6 months after grafting stained for human synaptophysin andSC121 in a pre-embedding fashion confirmed the presence ofsynaptic structures originating from differentiated human NPs.Human synaptophysin(#) terminals were especially densearound the graft, in substantia innominata and in the bed nu-cleus of stria terminalis. Labeled terminals were also evidentin frontal, motor, somatosensory, and auditory cortex and inthe olfactory bulb. Many labeled structures were en passantswellings with round (38–50 nm in diameter) vesicles, sugges-tive of synaptic vesicles, but did not exhibit pre- and postsy-naptic membrane specializations (not shown). Numerous la-beled terminal swellings were also observed. In a few ofthese terminals, synaptic vesicles and membrane specializa-tions were visible at the same plane, allowing for their unam-biguous classification as synapses (Fig. 7D–7F). In all thesesynapses, the unlabeled postsynaptic elements had the

Figure 6. Astroglial differentiation of HESC-derived NPs in fornix (A, B), genu of corpus callosum (C), and internal capsule/globus pallidus(C, D). Sections were processed for ICC with GFAP (A) or human cytosol-specific SC121 (B–D) antibodies. The section in (A) was processedwith immunofluorescence; sections in other panels were processed with immunoperoxidase. (A, B): In this case of spurious engraftment of humanNPs into the fornix bundle, the vast majority of HNu(#) or SC121(#) cells displayed astrocytic cytologies [both (A) and (B)] and GFAP expres-sion (A). Insets in (A) represent a confocal microscopic analysis of the profile indicated by connecting lines; top inset is a traditional confocalappearance of the index profile with a subsequent virtual resectioning at the x- and y-axis; bottom inset is a compressed z-stack image of theindex profile. Inset in (B) shows a low-power image of the fornix from which the main panel was derived. (C, D): In this case of a peripheralinvolvement of the callosal genu (C) or globus pallidus (D) in the graft, there was a predominant differentiation of human NPS into astrocytes.Insets in (C) and (D) are magnifications of framed areas in main panels. Scale bars: (A, C, D): 100 lm; (B): 10 lm.

2422 Differentiation of HESC-Derived Neural Precursors

structure of dendritic shafts or spines. Unlabeled postsynapticstructures were inferred to be of host (rat) origin due to theirdistance from the graft and the paucity of parenchymal migra-tion of differentiated NPs. SC121 immunoreactivity was very

similar in ultrastructural appearance to that of humansynaptophysin.

In concert, differentiated grafts of human NPs show anearly elaboration of axons that form bundles and traverse

Figure 7. Evidence for some local and remote integration of differentiated human NPs in rat forebrain circuitry manifested by early (A) and late (B)pathway formation, remote synaptic configurations (C), and the formation of mature synapses as evidenced by immuno-EM labeling for human synap-tophysin and the human cytosolic marker SC121 (D–F). (A): Early projections from the 1.5-month graft visualized by doublecortin (arrowheads indi-cate bundles of pioneer axons). Inset in (A), top, represents a magnification of the axonal front marked with an asterisk in main panel. Inset in (A),bottom, represents a confocal microscopic analysis, with a subsequent virtual resectioning at the x- and y-axis, of the framed area in main graft. (B):Extensive formation of axon bundles by mature (6-month old) differentiated grafts occurs along most established forebrain pathways including thestriato-pallidal pencils and the anterior commissure (ac and sppen, top left), thalamic striae and the mamillothalamic tract (tstriae and mtt, top right),the rostral migratory stream (RMS, bottom left), and the genu-forceps minor of the corpus callosum (fm/cc, bottom right). In this panel, graft-derivedaxon bundles are visualized with SC121 ICC. Sections are counterstained with fast green. (C): Synaptophysin immunoreactivity in terminal fields in6-month-old grafts. These panels illustrate representative synaptic boutons in the olfactory bulb (OB) (internal plexiform and outer granule cell layer),layers I and II of the frontal cortex (Cx), the globus pallidus (GP), and the islands of Calleja (IC). Insets represent magnifications of areas indicatedwith asterisks. (D–F): Electron microscopic images of synapses with human signature [(D, E): human synaptophysin labeling; (F): SC121 labeling]taken close to the striatal graft (D), in the nucleus accumbens (B), and in the glomerular layer of the olfactory bulb (C). Sections in (D) and (E) wereprepared 3 months after grafting; section in (F) was prepared 6 months after grafting. Note postsynaptic densities (arrows) and synaptic vesicles in alllabeled terminals. Abbreviations: ac, anterior commissure; cc, corpus callosum; fm, forceps minor of corpus callosum; GR, graft; mtt, mammillothala-mic tract; RMS, rostral migratory stream; sppen, striato-pallidal pencils; tstriae, thalamic striae. Scale bars: (A, B): 100 lm; (C): 10 lm.

Nasonkin, Mahairaki, Xu et al. 2423

www.StemCells.com

along established pathways in the host forebrain. Later graftsshow an abundance of synaptophysin- or SC121-immunoreac-tive boutons in terminal fields; a number of these boutonsdemonstrate ultrastructural features of mature synapses.

DISCUSSION

General ConclusionsOur findings show an extensive neuronal differentiation ofHESC-derived NPs and considerable structural integrationinto the adult rodent forebrain. The long time points used inthis paper (up to 6 months) serve to show the stable, tumor-free differentiation that is key for the safety and therapeuticpotential of HESC-derived NP grafts [30]. The long survivaltimes used here also show the ongoing differentiation of theseHESC-derived cells several months after grafting, that is, timeframes that are unusual in stem cell transplantation studies[31], but may be necessary for establishing the full graftingpotential of the slowly differentiating human cells. The num-ber of NPs used for transplantation appears to be an importantvariable that, when optimized to the appropriate level (15 !103 cells per animal), may eliminate neuroepithelial tube for-mation and safeguard the host from tumorigenesis and graftovergrowth [32].

NPs used in this paper have been derived by exposing ad-herent monolayers of HESCs to high concentrations of nogginwhich, by blocking BMP, arrests non-neural fate choices [33]and, further downstream, may also inhibit astroglial differen-tiation [34]. The pluripotent markers Oct3/4 and Nanog arealmost eliminated by day 28 of the noggin-based protocol dueto the efficient neuralization of HESCs, thus drastically reduc-ing the tumorigenic potential of grafted cells. Despite a pre-dominant neuronal differentiation, such derived NPs were ca-pable of making astroglial fate choices when inoculated intowhite matter tracks, a finding that supports the identity of aportion of these NPs as neural stem cells at the time of trans-plantation. Further support of their plastic nature at the timeof transplantation is the arrest of these NPs in a nestin(#)state in the case of inoculation in the ependymal region.

Further Comments on Culturing MethodologyThe culture conditions used here to prepare NPs from HESCsand expand them prior to grafting were established in a previ-ous report [20] in which we have also shown the karyotypicstability and lack of Neu5Gc contamination of these precur-sors. The present paper demonstrates further the advantage ofthe adherent monolayer system in effecting a synchronizeddownregulation of pluripotency genes such as Nanog andOct3/4 [35] and the upregulation of neural markers. Sox2, agenetic marker of both HESCs and neural stem cells [36], isexpressed throughout differentiation. P2 proved to be a goodtime choice for transplantation of cells treated with our nog-gin-based differentiation protocol. At this time point, cellsappear to be plastic enough to acquire either neuronal orastroglial fates in vivo, yet also mature enough to avoid ex-cessive proliferation. Similar differentiation patterns wereseen when the HESC lines BG02 TE03 ES04 were used inlieu of BG01 under the exact same culture conditions (I.Nasonkin, unpublished observations).

Most NPs at the time of grafting were early neuronal pro-genitors, although one fifth or so (the cell population exhibit-ing the CD15#TnTx% immunophenotyping profile) may ex-hibit a broader fate plasticity and generate both neurons andglial cells [37]. Another NP subpopulation defined by the

CD15%TnTx% CD29# profile may give rise to radial glia and/or astrocytes.

Optimizing between proliferation propensity and neural‘‘stemness’’ of precursors is a challenging task that has beenaddressed and managed before [38]. A prompt reduction ofmitotic index has been consistent across grafting experimentsand is an especially encouraging finding of the present study.Reduction in mitoses was eventually also encountered in pilotgrafts with much higher numbers of cells than the ones usedin the main study, despite the tendency of such grafts to gothrough an initial period of neuroepithelial proliferation (sup-porting information Fig. 1). In contrast to forebrain transplan-tation, NP grafting into the brainstem and spinal cord showeda propensity for tumorigenesis and was not pursued further.

Migration and Differentiation Choicesof Human NPsOn one hand, as we observed previously in the case of spinalcord grafts [2], the host microenvironment plays a decisiverole in determining differentiation choices of grafted NPs. Onthe other hand, the large population of neuronal progenitors inour NP preparation is certainly consistent with the predomi-nant neuronal differentiation of these NPs in the rat forebrain.Astrocytes present in mature grafts originated either in neuralstem cells or in the small number of astroglial precursorspresent in the inoculum. Besides cell-intrinsic signals, such asPOU, Notch, and bHLH, extrinsic molecular cues includingnoggin, BMPs, bFGF, EGF, LIF, and neurotrophins are likelyto play key roles in promoting differentiation choices faithfulto the cellular composition of each one of these diversemicroenvironments [39–44]. The neuronal bias in differentia-tion of striatal grafts was evident in vivo with the promptexpression of doublecortin and then TUJ1 and synaptophysinby 3 months after grafting, although the expression of neurofi-lament markers and axonal elongation/pathfinding and elabo-ration of terminals for most forebrain targets would takeanother 3 months. This long time course of in vivo differen-tiation may be appropriate for the longer differentiation cycleof human NPs [45] and argues for the extension of survivaltimes in human NP grafting experiments beyond the conven-tional 10–12 weeks. Despite advanced neuronal differentia-tion, including the late acquisition of DARPP-32 by one thirdof human NPs, these cells did not appear to achieve definitiveneurotransmitter specialization, that is, become GABAergicneurons.

DARPP-32 is an early marker of lateral ganglionic emi-nence (LGE)-derived basal forebrain neurons, including me-dium spiny striatal projection neurons [46], and has beenfound in at least one other occasion in neurons differentiatingfrom primate ES-derived progenitors grafted into striatum[47]. Expression of DARPP-32 in graft-derived neurons indi-cates that a substantial number of grafted NPs correctlyacquired basal forebrain identity, yet failed to achieve theexpected GABAergic phenotype of striatal projection neuronsby 6 months after grafting. It is possible that whereas striatalprogenitor cells acquire GABAergic and DARPP-32 identitiesunder the influence of synchronized inductive signals in thedeveloping LGE, in artificial systems combining initial invitro differentiation followed by further maturation in theadult striatal environment NPs may acquire the one (i.e.,DARPP-32) but not the other (i.e., GABA) differentiationmarker [46, 48–52]. The importance of temporally and posi-tionally appropriate inductive signaling is evident in studiesusing ex vivo derived NSCs as spinal cord grafts [1, 2], wherespinal cord-derived NPs may be better able to undergo neuro-nal differentiation to GABAergic neurons compared to brain-

2424 Differentiation of HESC-Derived Neural Precursors

derived precursors [53]. Another possibility is that humanNPs require more time in vivo to achieve full GABAergicdifferentiation.

Axons, Synapses, and Structural Integrationin the Host ForebrainSC121 ICC labels all axons projecting from differentiatedNPs and shows that these axons take established pathways intheir vicinity but do not grow through the gray matter; indoing so, these axons reach both the appropriate innervationtargets of the striatum such as globus pallidus and subthala-mic nucleus but also nonconstitutive targets such as neocor-tex. This pattern is consistent with previous findings thatintact white matter tracts organize the axonal trajectories oftransplanted developing [54, 55] and adult [56] neurons in aparallel molding fashion that is mediated, among other things,by a neurite-inhibitory effect of myelin [57].

The synaptic nature of terminal or en passant swellings la-beled with human synaptophysin and SC121 was ascertainedhere by the presence of synaptic vesicles accumulating nextto dense material on the cytoplasmic side of the terminalmembrane [58], a synaptic cleft 20–30-nm wide, and an accu-mulation of dense cytoplasmic material on the postsynapticmembrane. The identity of such profiles as synapses betweengraft-derived and host neurons was based on either a combi-nation of human synaptophysin immunoreactivity of the syn-apse and a postsynaptic structure remote from the graft or thecombination of SC121 immunoreactivity of the synapse andan unlabeled postsynaptic structure in any location. Synapticspecializations between graft and host neurons were not visi-ble at 3 months but were apparent at 6 months, that is, tookseveral months before these human neurons established synap-tic contacts with host postsynaptic structures.

Relevance for Experimental TherapeuticsOur findings demonstrate the generic ability of NPs derivedfrom noggin-treated human HESCs and optimized for cellnumber and maturation stage, to survive and differentiate inthe adult CNS without evident tumorigenesis. Despite a pre-dominantly neuronal bias of grafted NPs based on the culture

conditions used here, these cells maintain their cell fate plas-ticity and continue to respond to differentiation cues from par-ticular CNS microenvironments.

The substantial degree of neuronal differentiation of ourstriatal NP grafts raises the issue of their potential as replace-ment therapies for neurodegenerative diseases of the basal gan-glia, such as Huntington’s disease. Such enthusiasm, however,should be tempered by the fact that these cells do not showcomplete differentiation into the main projection neurons ofthe striatum, that is, medium spiny neurons. With progress inour understanding of molecular cues specifying developmentalfate choices in the basal ganglia-basal forebrain, one can fore-see modifications of the in vitro protocol reported here orongoing treatment of these grafts in vivo. These additions willhopefully optimize the differentiation of human NPs and deter-mine their potential as cell replacement therapies for diseasesof the basal ganglia. In the mean time, their experimental thera-peutic value lies primarily in their ability to make synaptic con-tacts with host neurons and, via these contacts, support injuredhost neurons with released trophic signals [41].

ACKNOWLEDGMENTS

We thank Alex Kecojevic, Annie Welsh, Tan Pongstaporn, andChrista Baker for their technical help and StemCells Inc. and Dr.Christopher Walsh for their generous gifts of the SC121 andDCAMKL1 antibodies, respectively. This work was supportedby the National Institutes of Health grants NS45140-03, NIDCDDC000232, NIDCD P30 DC005211, and EY01765, the RobertPackard Center for ALS Research at Johns Hopkins, and theMuscular Dystrophy Association.

DISCLOSURE OF POTENTIAL CONFLICTS

OF INTEREST

The authors indicate no potential conflicts of interest.

REFERENCES

1 Cummings BJ, Uchida N, Tamaki SJ et al. Human neural stem cellsdifferentiate and promote locomotor recovery in spinal cord-injuredmice. Proc Natl Acad Sci U S A 2005;102(39):14069–14074.

2 Yan J, Xu L, Welsh AM et al. Extensive neuronal differentiation ofhuman neural stem cell grafts in adult rat spinal cord. PLoS Med2007;4(2):e39.

3 Tabar V, Panagiotakos G, Greenberg ED et al. Migration and differen-tiation of neural precursors derived from human embryonic stem cellsin the rat brain. Nat Biotechnol 2005;23(5):601–606.

4 Takahashi K, Tanabe K, Ohnuki M et al. Induction of pluripotentstem cells from adult human fibroblasts by defined factors. Cell 2007;131(5):861–872.

5 Mitalipova M, Calhoun J, Shin S et al. Human embryonic stem celllines derived from discarded embryos. Stem Cells 2003;21(5):521–526.

6 Perrier AL, Tabar V, Barberi T et al. Derivation of midbrain dopa-mine neurons from human embryonic stem cells. Proc Natl Acad SciU S A 2004;101(34):12543–12548.

7 Chung S, Shin BS, Hwang M et al. Neural precursors derived fromembryonic stem cells, but not those from fetal ventral mesencephalon,maintain the potential to differentiate into dopaminergic neurons afterexpansion in vitro. Stem Cells 2006;24(6):1583–1593.

8 Zeng X, Cai J, Chen J et al. Dopaminergic differentiation of humanembryonic stem cells. Stem Cells 2004;22(6):925–940.

9 Shin S, Dalton S, Stice SL. Human motor neuron differentiation fromhuman embryonic stem cells. Stem Cells Dev 2005;14(3):266–269.

10 Lee H, Shamy GA, Elkabetz Y et al. Directed differentiation andtransplantation of human embryonic stem cell-derived motoneurons.Stem Cells 2007;25(8):1931–1939.

11 Li XJ, Du ZW, Zarnowska ED et al. Specification of motoneurons fromhuman embryonic stem cells. Nat Biotechnol 2005;23(2):215–221.

12 Keirstead HS, Nistor G, Bernal G et al. Human embryonic stem cell-derived oligodendrocyte progenitor cell transplants remyelinate andrestore locomotion after spinal cord injury. J Neurosci 2005;25(19):4694–4705.

13 Bain G, Kitchens D, Yao M et al. Embryonic stem cells express neu-ronal properties in vitro. Dev Biol 1995;168(2):342–357.

14 Okabe S, Forsberg-Nilsson K, Spiro AC et al. Development of neuro-nal precursor cells and functional postmitotic neurons from embryonicstem cells in vitro. Mech Dev 1996;59(1):89–102.

15 Schulz TC, Palmarini GM, Noggle SA et al. Directed neuronal differ-entiation of human embryonic stem cells. BMC Neurosci 2003;4:27.

16 Zhang SC, Wernig M, Duncan ID et al. In vitro differentiation oftransplantable neural precursors from human embryonic stem cells.Nat Biotechnol 2001;19(12):1129–1133.

17 Shin S, Mitalipova M, Noggle S et al. Long-term proliferation ofhuman embryonic stem cell-derived neuroepithelial cells using definedadherent culture conditions. Stem Cells 2006;24(1):125–138.

18 McMahon JA, Takada S, Zimmerman LB et al. Noggin-mediated an-tagonism of BMP signaling is required for growth and patterning ofthe neural tube and somite. Genes Dev 1998;12(10):1438–1452.

19 Gerrard L, Rodgers L, Cui W. Differentiation of human embryonicstem cells to neural lineages in adherent culture by blocking bonemorphogenetic protein signaling. Stem Cells 2005;23(9):1234–1241.

20 Nasonkin IO, Koliatsos VE. Nonhuman sialic acid Neu5Gc is verylow in human embryonic stem cell-derived neural precursors

Nasonkin, Mahairaki, Xu et al. 2425

www.StemCells.com

differentiated with B27/N2 and noggin: Implications for transplanta-tion. Exp Neurol 2006;201(2):525–529.

21 Maric D, Barker JL. Fluorescence-based sorting of neural stem cellsand progenitors. Curr Protoc Neurosci 2005;Chapter 3:Unit 3.18.

22 Xu L, Yan J, Chen D et al. Human neural stem cell grafts amelioratemotor neuron disease in SOD-1 transgenic rats. Transplantation 2006;82(7):865–875.

23 Maric D, Maric I, Chang YH et al. Stereotypical physiological proper-ties emerge during early neuronal and glial lineage development inthe embryonic rat neocortex. Cereb Cortex 2000;10(8):729–747.

24 Maric D, Fiorio Pla A, Chang YH et al. Self-renewing and differenti-ating properties of cortical neural stem cells are selectively regulatedby basic fibroblast growth factor (FGF) signaling via specific FGFreceptors. J Neurosci 2007;27(8):1836–1852.

25 Przyborski SA. Differentiation of human embryonic stem cells aftertransplantation in immune-deficient mice. Stem Cells 2005;23(9):1242–1250.

26 Cooke MJ, Stojkovic M, Przyborski SA. Growth of teratomas derivedfrom human pluripotent stem cells is influenced by the graft site.Stem Cells Dev 2006;15(2):254–259.

27 Eberhart CG. In search of the medulloblast: Neural stem cells andembryonal brain tumors. Neurosurg Clin N Am 2007;18(1):59–69,viii–ix.

28 Deuel TA, Liu JS, Corbo JC et al. Genetic interactions between dou-blecortin and doublecortin-like kinase in neuronal migration and axonoutgrowth. Neuron 2006;49(1):41–53.

29 Koizumi H, Tanaka T, Gleeson JG. Doublecortin-like kinase functionswith doublecortin to mediate fiber tract decussation and neuronalmigration. Neuron 2006;49(1):55–66.

30 Isacson O. The production and use of cells as therapeutic agents inneurodegenerative diseases. Lancet Neurol 2003;2(7):417–424.

31 Sonntag KC, Pruszak J, Yoshizaki T et al. Enhanced yield of neuroe-pithelial precursors and midbrain-like dopaminergic neurons fromhuman embryonic stem cells using the bone morphogenic protein an-tagonist noggin. Stem Cells 2007;25(2):411–418.

32 Aubry L, Bugi A, Lefort N et al. Striatal progenitors derived fromhuman ES cells mature into DARPP32 neurons in vitro and in quino-linic acid-lesioned rats. Proc Natl Acad Sci U S A 2008;105(43):16707–16712.

33 Munoz-Sanjuan I, Brivanlou AH. Neural induction, the default modeland embryonic stem cells. Nat Rev Neurosci 2002;3(4):271–280.

34 Rajan P, Panchision DM, Newell LF et al. BMPs signal alternatelythrough a SMAD or FRAP-STAT pathway to regulate fate choice inCNS stem cells. J Cell Biol 2003;161(5):911–921.

35 Boyer LA, Mathur D, Jaenisch R. Molecular control of pluripotency.Curr Opin Genet Dev 2006;16(5):455–462.

36 Kuroda T, Tada M, Kubota H et al. Octamer and Sox elements arerequired for transcriptional cis regulation of Nanog gene expression.Mol Cell Biol 2005;25(6):2475–2485.

37 Capela A, Temple S. LeX/ssea-1 is expressed by adult mouse CNSstem cells, identifying them as nonependymal. Neuron 2002;35(5):865–875.

38 Brederlau A, Correia AS, Anisimov SV et al. Transplantation ofhuman embryonic stem cell-derived cells to a rat model of Parkin-son’s disease: Effect of in vitro differentiation on graft survival andteratoma formation. Stem Cells 2006;24(6):1433–1440.

39 Tomita K, Moriyoshi K, Nakanishi S et al. Mammalian achaete-scuteand atonal homologs regulate neuronal versus glial fate determinationin the central nervous system. EMBO J 2000;19(20):5460–5472.

40 Chen H, Tung YC, Li B et al. Trophic factors counteract elevatedFGF-2-induced inhibition of adult neurogenesis. Neurobiol Aging2007;28(8):1148–1162.

41 Lim DA, Tramontin AD, Trevejo JM et al. Noggin antagonizes BMPsignaling to create a niche for adult neurogenesis. Neuron 2000;28(3):713–726.

42 Weible MW II, Chan-Ling T. Phenotypic characterization of neuralstem cells from human fetal spinal cord: Synergistic effect of LIF andBMP4 to generate astrocytes. Glia 2007;55(11):1156–1168.

43 Castro DS, Skowronska-Krawczyk D, Armant O et al. ProneuralbHLH and Brn proteins coregulate a neurogenic program through co-operative binding to a conserved DNA motif. Dev Cell 2006;11(6):831–844.

44 Ge W, He F, Kim KJ et al. Coupling of cell migration with neurogen-esis by proneural bHLH factors. Proc Natl Acad Sci U S A 2006;103(5):1319–1324.

45 Naimi S, Jeny R, Hantraye P et al. Ontogeny of human striatalDARPP-32 neurons in fetuses and following xenografting to the adultrat brain. Exp Neurol 1996;137(1):15–25.

46 Deacon TW, Pakzaban P, Isacson O. The lateral ganglionic eminence isthe origin of cells committed to striatal phenotypes: Neural transplanta-tion and developmental evidence. Brain Res 1994;668(1–2):211–219.

47 Ferrari D, Sanchez-Pernaute R, Lee H et al. Transplanted dopamineneurons derived from primate ES cells preferentially innervateDARPP-32 striatal progenitors within the graft. Eur J Neurosci 2006;24(7):1885–1896.

48 Lauder JM, Han VK, Henderson P et al. Prenatal ontogeny of theGABAergic system in the rat brain: An immunocytochemical study.Neuroscience 1986;19(2):465–493.

49 Stenman J, Toresson H, Campbell K. Identification of two distinctprogenitor populations in the lateral ganglionic eminence: Implicationsfor striatal and olfactory bulb neurogenesis. J Neurosci 2003;23(1):167–174.

50 Allain AE, Meyrand P, Branchereau P. Ontogenic changes of the spi-nal GABAergic cell population are controlled by the serotonin (5-HT)system: Implication of 5-HT1 receptor family. J Neurosci 2005;25(38):8714–8724.

51 Wichterle H, Turnbull DH, Nery S et al. In utero fate mapping revealsdistinct migratory pathways and fates of neurons born in the mamma-lian basal forebrain. Development 2001;128(19):3759–3771.

52 Lin PY, Hinterneder JM, Rollor SR et al. Non-cell-autonomous regu-lation of GABAergic neuron development by neurotrophins and thep75 receptor. J Neurosci 2007;27(47):12787–12796.

53 Koliatsos VE, Xu L, Yan J. Human stem cell grafts as therapies formotor neuron disease. Expert Opin Biol Ther 2008;8(2):137–141.

54 Wictorin K, Brundin P, Gustavii B et al. Reformation of long axonpathways in adult rat central nervous system by human forebrain neu-roblasts. Nature 1990;347(6293):556–558.

55 Wictorin K, Clarke DJ, Bolam JP et al. Extensive efferent projectionsof intra-striatally transplanted striatal neurons as revealed by a spe-cies-specific neurofilament marker and anterograde axonal tracing.Prog Brain Res 1990;82:391–399.

56 Davies SJ, Fitch MT, Memberg SP et al. Regeneration of adult axonsin white matter tracts of the central nervous system. Nature 1997;390(6661):680–683.

57 Pettigrew DB, Crutcher KA. Myelin contributes to the parallel orienta-tion of axonal growth on white matter in vitro. BMC Neurosci 2001;2:9.

58 Schikorski T, Stevens CF. Quantitative ultrastructural analysis of hip-pocampal excitatory synapses. J Neurosci 1997;17(15):5858–5867.

Seewww.StemCells.com for supporting information available online.

2426 Differentiation of HESC-Derived Neural Precursors