TGF-β family signaling in stem cells

Biochimica et Biophysica Acta 1830 (2013) 2280–2296

Contents lists available at SciVerse ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r .com/ locate /bbagen

Review

TGF-β family signaling in stem cells☆

Masayo Sakaki-Yumoto, Yoko Katsuno, Rik Derynck ⁎Department of Cell and Tissue Biology, University of California at San Francisco, San Francisco, CA 94143-0669, USADepartment of Anatomy, University of California at San Francisco, San Francisco, CA 94143-0669, USAEli and Edythe Broad Center of Regeneration Medicine and Stem Cell Research, University of California at San Francisco, San Francisco, CA 94143-0669, USA

☆ This article is part of a Special Issue entitled Bioche⁎ Corresponding author at: Department of Cell and

California at San Francisco, Dolby Regeneration MedicineMedical Center Way, USA. Tel.: +1 415 476 7322; fax:

E-mail address: [email protected] (R. Derynck).

0304-4165/$ – see front matter © 2012 Elsevier B.V. Alhttp://dx.doi.org/10.1016/j.bbagen.2012.08.008

a b s t r a c t
a r t i c l e i n f o
Article history:
Received 16 April 2012Received in revised form 11 July 2012Accepted 7 August 2012Available online 16 August 2012
Keywords:PluripotencySomatic stem cellReprogrammingCancer stem cellEpithelial–mesenchymal transitionNiche

Background: The diversity of cell types and tissue types that originate throughout development derives from thedifferentiation potential of embryonic stem cells and somatic stem cells. While the former are pluripotent, andthus can give rise to a full differentiation spectrum, the latter have limited differentiation potential but drivetissue remodeling. Additionally cancer tissues also have a small population of self-renewing cells with stem cellproperties. These cancer stem cells may arise through dedifferentiation from non-stem cells in cancer tissues,illustrating their plasticity, and may greatly contribute to the resistance of cancers to chemotherapies.Scope of review: The capacity of the different types of stem cells for self-renewal, the establishment andmaintenance of their differentiation potential, and the selection of differentiation programs are greatly definedby the interplay of signaling molecules provided by both the stem cells themselves, and their microenviron-ment, the niche. Here we discuss common and divergent roles of TGF-β family signaling in the regulation ofembryonic, reprogrammed pluripotent, somatic, and cancer stem cells.Major conclusions: Increasing evidence highlights the similarities between responses of normal and cancer stem
cells to signaling molecules, provided or activated by their microenvironment. While TGF-β family signalingregulates stemness of normal and cancer stem cells, its effects are diverse and depend on the cell types andphysiological state of the cells.General significance: Further mechanistic studies will provide a better understanding of the roles of TGF-β familysignaling in the regulation of stem cells. These basic studies may lead to the development of a new therapeuticor prognostic strategies for the treatment of cancers. This article is part of a Special Issue entitled Biochemistryof Stem Cells.
© 2012 Elsevier B.V. All rights reserved.

1. Introduction

Stem cells are undifferentiated cells that have an indefiniteexpansion potential to produce progeny through self-renewal ordifferentiation processes. They exist in embryonic tissues, as well as inpostnatal and adult tissues. The stem cells that have received mostvisibility are the pluripotent embryonic stem cells (ESCs), which arederived from the inner cell mass of blastocyst stage embryos and giverise to all three germ layers (Fig. 1) [1–4]. Other pluripotent stem cellsexist, such as the epiblast stem cells (EpiSCs), which are originallyderived from the epiblast of mouse post implantation stage (E5.5–6.5)embryos and regarded as cells that are more similar to human thanmouse ESCs [5,6]. Following early embryogenesis, most organs haveresidentmultipotent stemcells that can give rise to amore limited set oflineages. These are called somatic or tissue stem cells (Fig. 1). Thesestem cells multiply through symmetric or asymmetric cell divisions to

mistry of Stem Cells.Tissue Biology, University ofBuilding, Room RMB 1027, 35

+1 415 502 7338.

l rights reserved.

give rise to new stem cells as well as differentiated cell types, replenishdying cells and regenerate damaged tissues. Due to the inherentdifferentiation plasticity of stem cells, extrinsic growth and differenti-ation factors, or ectopically expressed key transcription factors, have theability to direct or redirect differentiation (Fig. 1) [7–10]. Through theorchestrated balance of self-renewal and differentiation, tissuesmaintain their homeostasis. The properties of both ESCs and somaticstem cells are determined and maintained by their local cell environ-ment, i.e. the stem cell niche [11–14]. For example, stem cell nichesmaintain somatic stem cells in quiescence, but, after tissue injury, themicroenvironment signals stem cells to promote either self-renewal ordifferentiation to form new tissues. The niche saves stem cells fromdepletion, while protecting the host from abnormal stem cell prolifer-ation. The interplay between stem cells and their niche creates adynamic system required to sustain tissue integrity.

In addition to these natural stem cells, induced pluripotent stem(iPS) cells have been derived from differentiated cells. By transientlyexpressing the transcription factors Oct4, Sox2, Klf4 and c-Myc, or somealternative ones, differentiated adult cells, such as fibroblasts or skincells, can be reprogrammed to dedifferentiate into pluripotent cells thatacquire many characteristics of ESCs [15–17] (Fig. 1). This paradigm-shifting technique of somatic cell reprogramming has facilitated the

http://dx.doi.org/10.1016/j.bbagen.2012.08.008

mailto:[email protected]

http://dx.doi.org/10.1016/j.bbagen.2012.08.008

http://www.sciencedirect.com/science/journal/03044165

Fertilization

Somatic cells

8 cell embryo Blastocyst

Embryonic Stem CellsiPS cells

SomaticStem Cells

MelanocyteSkin stem cells

Hematopoietic stem cells

Mammary stem cells

Neural stem cells

Reprogramming

Differentiation

Neuron Isolation

ICM

TE

Feeder cells

Cancer Stem Cells

Stromal cell

Cancer stem cells

Somatic stem cell

Oncogenic change

Cancer associated fibroblast

Tumor

Normal niche

Cancer stem cell niche

Epithelial tissue

Fig. 1. Schematic figure of four types of stem cells. Yellow area. Embryonic stem cells are derived from inner cell mass (ICM) of blastocyst stage embryos (green arrow). Inducedpluripotent stem (iPS) cells are generated by reprogramming somatic cells into pluripotent cells (orange arrow). TE: trophectoderm. Pink area. Tissue stem cells, which reside inadult tissues, are the source of functionally committed cells. They generate all cell types in the same tissue through differentiation (red arrow). Some tissue stem cells are known totransdifferentiate toward other lineages (not shown). Blue area. Cancer stem cells (CSCs) represent a small subpopulation of undifferentiated cancer cells that have stem cell-likeproperties and are responsible for tumor expansion. The normal epithelial stem cell niche contains stromal cells with a mesenchymal phenotype that support tissue homeostasis.Oncogenic changes (red arrow) in normal somatic stem cells initiate tumorigenesis. CSCs and tumor tissues affect the properties of the stromal cells constituting the normal niche.The altered niche, where tumor stromal cells, including cancer-associated fibroblasts (CAFs), are recruited, supports the maintenance of CSCs.

2281M. Sakaki-Yumoto et al. / Biochimica et Biophysica Acta 1830 (2013) 2280–2296

generation, by directly reprogramming somatic cells with lineage-specific master genes, of a variety of differentiated cell types withdefined functions, such as pancreatic β cells, brown adipocytes, cardiacmyoblasts, natural killer cells and neurons [18–22].

A rare population of cells in cancer tissues has also been shown tohave stem cell properties, including self-renewal capacity and multi-potency, and is required for tumor formation and maintenance. Thissmall fraction of cancer cells, so-called cancer stem cells (CSCs), hasbeen identified in tumors of various organs and tissues [23,24].Although the origin of the CSCs is subject of debate, an increasingnumber of studies indicate that many types of tumors initiate fromnormal tissue stem cells that acquired oncogenic mutations, and notfrom differentiated cells [25] (Fig. 1). Similarly to the normal stem cellniches, CSCs are supported by functional local microenvironments,and reciprocal interactions exist between CSCs and their microenvi-ronment. CSCs affect the properties of the adjacent stromal cells andseem to alter the normal stem cell niche, while signals from anaberrant niche that mimics the normal stem cell niche helps maintainCSC properties [24,26] (Fig. 1).

Signaling by TGF-β family ligands plays key roles in cell differenti-ation and proliferation, and is important for many stem cell types.Among the TGF-β family proteins, signaling by TGF-β or activin proteinsis essential formaintaining pluripotency of human ESCs [27] andmouseEpiSCs, and helps define the differentiation potential and proliferationof these cell types. Also somatic stem cells rely on TGF-β familysignaling, either provided in an autocrine fashion, or by factors in theirmicroenvironment where they reside. For example, quiescent hair

follicle stem cells are activated for regeneration by TGF-β that isprovided by the underlying mesenchymal dermal papillae [28]. TGF-βexpressed by the follicle stem cells is also known to serve as a nichesignal for melanocyte stem cells to enter quiescence [29]. Similarly tonormal stem cells in development, TGF-β family signaling from themicroenvironment also regulates the properties of the cancer stem cellpopulation. One of the key events in acquiring stem cell properties ofboth breast cancer and normal mammary stem cells is an epithelial-mesenchymal transition (EMT) induced by TGF-β [30,31]. While TGF-βfamily signaling regulates stemness of normal and neoplastic stem cells,its effects are diverse and depend on the cell types, and microenviron-ment and physiological state of the cells. This review will focus on theroles of TGF-β family signaling in diverse functions and properties ofESCs, iPS cells, somatic stem cells, and CSCs.

2. TGF-β family signaling

The TGF-β family is encoded by 33 genes encoding structurallyrelated polypeptides that correspond to ligand precursors. Thesecomprise a large propeptide and the C-terminal mature polypeptide thatis proteolytically cleaved from the precursor [32]. TGF-β family ligandsare disulfide-linked homodimers or heterodimers of the C-terminalpolypeptides, and include TGF-βs, activins, nodal, “growth and differen-tiation factors” (GDFs) and bone morphogenetic proteins (BMPs)[33,34]. TGF-β and myostatin are secreted as complexes with otherproteins that prevent ligand binding to the receptors. These latent TGF-βcomplexes are activated and released by proteolytic cleavage of the

2282 M. Sakaki-Yumoto et al. / Biochimica et Biophysica Acta 1830 (2013) 2280–2296

propeptide [32]. TGF-β familymembers regulate a wide range of cellularresponses, including most if not all cell differentiation events [35–38].

Signaling mediated by TGF-β ligands is transduced through cellsurface receptor complexes of two distinct types of transmembranekinases, the type I and type II receptors [34,39] (Fig. 2). In mammals,seven type I receptors (ALK‐1–7) and five type II receptors exist, and avariety of heteromeric combinations have been defined. Ligandbinding stabilizes the formation of a tetrameric receptor complex,consisting of a pair of type II and a pair of type I receptors, in whichthe type II receptors, with their constitutively active kinases,phosphorylate the type I receptors. TGF-β and activin primarilyactivate the type I receptors TβRI/ALK-5 and ActRIB/ALK-4, respec-tively, whereas BMPs activate ALK-1, ALK-2, BMPRIA/ALK-3 andBMPRIB/ALK-6. Nodal activates ActRIB/ALK-4 and ALK-7. The type Ireceptors are then poised to activate effector Smads and othersignaling mediators that drive non-Smad signaling pathways [34,39](Fig. 2). Activation of either Smad or non-Smad pathways by aspecific ligand depends on many factors, including cell surfacereceptors, co-receptors, antagonists and intracellular co-factors.Differential expression of these regulators in stem cells and theniche results in the diverse effects of TGF-β family signaling ondifferent stem cells.

2.1. Smad-mediated signaling

In vertebrates, there are eight Smads, Smad1 to Smad8 [40]. Uponligand stimulation and subsequent activation by type II receptors,type I receptors transmit intracellular signaling through phosphory-lation of downstream effector Smads [40–42]. Smad1, Smad5 andSmad8 are activated by BMP receptors, whereas Smad2 and Smad3are activated by TGF-β/activin/nodal receptors. These receptor-activated Smads (R-Smads) form trimers with Smad4, the commonSmad shared by the TGF-β/activin/nodal and BMP signaling path-ways, and translocate into the nucleus (Fig. 2). The R-Smads, exceptfor Smad2, can bind to preferred DNA sequences, yet their affinity isnot sufficient to support the association of Smads with regulatory

PP

Smad4

R-Smad

P P

P PPP

Smad comp

transcription factor

transcriptioncofactor

type II Rtype I R(ALK)

cytoplasm

nucleus

ligands

Smad pathway

Fig. 2. Smad-dependent and Smad-independent TGF-β family signaling. Ligands of TGF-βreceptors phosphorylate the type I receptors, which then phosphorylate and activate effecnucleus. The Smad complex interacts with other transcription factors, co-activators or co-resignaling cascades independent of Smad pathways. TGF-β activates the Ras–Raf–MEK–Erksignaling through activation of TAK1 by the TRAF6. TGF-β also activates the small GTPases

promoter sequences. Thus, Smad complexes interact and cooperatewith DNA sequence-specific transcription factors, and, together, theyregulate gene transcription with the help of coactivators andcorepressors [43]. In the Smad complexes, Smad4 acts as coactivatorof Smad-mediated transcription regulation. Many genes are activatedin response to TGF-β family ligands, whereas others are transcrip-tionally repressed. Various families of DNA-binding transcriptionfactors with which Smads can functionally interact have beenidentified. Moreover, coactivators, such as CBP and p300, or co-repressors, such as histone deacetylases, that interact with Smads,define whether the target genes are activated or repressed, and theextent of activation or repression [40,42]. Many of the responses ofstem cells to TGF-β family ligands are regulated by Smad-mediatedtranscription activation or repression of key genes. In various celltypes, and in embryonic stem cells, Smads cooperate with masterregulators of cell differentiation or pluripotency [44–53]. Moreover,Smads can regulate gene expression by initially using histone marksas a platform to switch master regulator genes from the poised to theactive state, and by recruiting histone demethylases [54,55].

2.2. Smad-independent signaling

TGF-β also activates, in a Smad-independent manner, signalingpathways that are generally considered as important effectorpathways for tyrosine kinase receptors [41,56,57]. TGF-β inducesactivation of these non-Smad pathways through interactions ofsignaling mediators with the type I or type II receptors, either directlyor through adaptor proteins. Depending on the cell system, thesenon-Smad pathways can also be activated indirectly as a consequenceof Smad-mediated changes in gene expression. TGF-β has beenshown to directly activate the Ras–Raf–MEK–Erk MAPK pathwaythrough association of ShcA with the TGF-β receptor complex, anddirect tyrosine phosphorylation of ShcA by TGF-β type I receptor inresponse to TGF-β, taking advantage of the dual specificity of the TGF-βreceptor kinases. The phosphorylated tyrosines on ShcA then provide adocking site for the recruitment of Grb2 and Sos, and this complex

lex

target genes

non-Smad pathways

TRAF6

TAK1

p38 JNK

PI3K

RhoAAkt

mTOR

Shc

Ras

Erk1/2

RafRac

Cdc42

cell responses

GTPases

family members bind to type I and type II receptors. Upon ligand binding, the type IItor Smads. The activated Smads form complexes with Smad4, and translocate into thepressors to regulate transcription of target genes. TGF-β also elicits activation of otherMAPK pathway through tyrosine phosphorylation of ShcA, and p38 and JNK MAPK

Rho, Rac and Cdc42, and the PI3K–Akt pathway.


initiates Ras activation leading to ErkMAPK signaling cascade [58]. TGF-β also induces p38 and JNK MAPK signaling through activation of TAK1by the ubiquitin ligase TRAF6 that interacts with the TGF-β receptorcomplex [59,60]. Furthermore, TGF-β also regulates the activities of thesmall GTPase proteins Rho, Rac and Cdc42, which regulate thecytoskeletal organization and gene expression [61–63], but howreceptor activation leads to signaling by small GTPases remains to bebetter defined. TGF-β-activated RhoA induces activation of its down-stream targets ROCK and LIM kinase [64]. Finally, TGF-β induces Aktactivation through PI3K [65,66]. Once activated, Akt initiates signalingpathways, e.g. through mTOR, that play roles in cell survival, growth,migration and invasion [67,68]. The roles of TGF-β-induced, Smad-independent signaling in stem cells are still unclear and remain to beelucidated.

3. ESCs and iPS cells

The first ESCs were isolated about 30 years ago from the inner cellmass of mouse blastocysts [1]. Nowadays, studies on mouse andhuman ESCs have revealed similarities, yet also differences in theproperties and behavior of ESCs from both species [1,3]. ESCs canexpand indefinitely in cell culture by maintaining themselves, aprocess named self-renewal, and produce most cell types thatconstitute the full organism, through various differentiation pro-grams. These two properties, the potential to self-renew and thepluripotency, are defining characteristics of ESCs.

iPS cells are pluripotent stem cells that are artificially derived fromnon-pluripotent, differentiated cells [15]. By ectopically expressingcertain genes, i.e. Oct4, Sox2, Klf4 and c-Myc, mouse and humansomatic cells have been shown to acquire pluripotency [15–17].iPS cells are similar to ESCs in many aspects, such as gene expression,chromatin modification patterns, cell doubling time, capacity of differ-entiation, chimera formation and germline transmission [15–17,69].However, they retain an epigeneticmemory of their somatic cell of origin[70–72], requiring further assessment of the full extent of their relation toESCs [73].

3.1. Maintenance of pluripotency

Several factors are commonly important in the maintenance ofpluripotency of ESCs and iPS cells, in both mouse and human cellsystems. However, mouse and human cell systems also havecharacteristics that distinguish both species [74]. For example, theircolony morphologies differ; mouse ESCs grow into compact, dome-shaped colonies, and are amenable to single-cell dissociation,whereas human ESCs form flat, two-dimensional colonies, and poorlysurvive single cell passaging. Most distinctly, in cell culture, mouseESCs require leukemia inhibitory factor (LIF) (Fig. 3, orange zone) andBMP4 to maintain pluripotency, while human ESCs are maintained inthe presence of basic fibroblast growth factor (bFGF) (Fig. 3, beigezone) and activin or TGF-β [27,75] (Fig. 3). In human cells, LIFsignaling is dispensable for the maintenance of ESCs, whereas BMPsignaling induces mesodermal and trophectodermal differentiation,and thus negatively affects cell pluripotency [5,6,76–78]. In mouseESCs, BMP4 induces expression of Id proteins, which are DNA-bindingtranscription factors that inhibit differentiation into neuronal line-ages [79,80]. The positive effect of BMP4 on self-renewal of mouseESCs is additionally accomplished by inhibition of both the Erk andp38 MAPK pathways [81]. In contrast to mouse ESCs, human ESCs andmouse EpiSCs rely on FGF-induced Erk MAPK signaling to maintainpluripotency and suppress apoptosis and neuronal differentiation.

Mouse EpiSCs can give rise to multiple lineages in vitro andteratomas in vivo. However, in contrast to mouse ESCs, these EpiSCshave a limited potential to contribute to the generation of chimeraswhen introduced into early embryos, and have not been reported toundergo germ line transmission [6,82]. Interestingly, mouse EpiSCs rely

on the same signaling pathways as human ESCs to maintain pluripo-tency (Fig. 3). Inhibition of TGF-β/activin signaling or activation of BMPsignaling results in loss of pluripotency of both human ESCs and mouseEpiSCs, while activation of TGF-β/activin signaling induces differentia-tion of mouse ESCs in the absence of LIF (Fig. 3). Although themaintenance of mouse ESC pluripotency does not rely on TGF-β/activinsignaling, it plays a key role in the proliferation of mouse ESCs [83].

It has been demonstrated that EpiSCs can be converted into mouseESC-like cells in the presence of LIF without introducing an exogenoustranscription factor, albeit with a lower efficiency [84]. This conversion,along with the withdrawal of TGF-β/activin and bFGF, can be enhancedby inhibiting the kinase activities of the TGF-β/activin type I receptors,thus inhibiting TGF-β/activin signaling, and/or inhibiting the Erk MAPKpathway, glycogen synthase kinase GSK3, or histone demethylase LSD1[75,85–88]. Some reports indicate that a similar conversion can beachieved in human ESCs as well [89,90].

TGF-β/activin signaling also regulates Nanog expression and in thisway acts to maintain pluripotency, in both mouse EpiSCs and humanESCs. Phosphorylated Smad2 or Smad3, which was not distinguishedwith the antibody used, was shown to bind to the Nanog proximalpromoter region, and thus direct Nanog expression [50,51]. Upon BMPsignaling, Nanog can interact with Smad1 and block BMP/Smadsignaling [49], and conversely Smad1 may attenuate Nanog's functionas pluripotency transcription factor. Nanog expression is heterogeneousin mouse or human pluripotent stem cells [91]. ESC markers, such asStella, which may be involved in chromosomal organization and RNAprocessing, the transcription factor Oct4, and the cell surface proteinSSEA3, also have revealed distinct subpopulations of mouse ESCs,EpiSCs, and substantial heterogeneity in human ESCs [92–94]. Since theresponse of stem cells to TGF-β/activin signaling may vary betweensubpopulations, the heterogeneous Nanog expression in ESCs mayoriginate from intrinsic heterogeneity in cell signaling of the cellpopulation. Further mechanistic studies will provide a deeper under-standing of the roles of TGF-β/activin signaling, in addition to its role inNanog gene regulation.

3.2. Differentiation

Since diverse and detailed effects of TGF-β family signaling on thedifferentiation of ESCs have been reported [95–100], we focus onthe presumed roles of TGF-β proteins in the very early lineagedecisions of ESCs. The formation of the primitive streak, whichprovides the origin of mesodermal and endodermal cells, representsthe first step of ESC differentiation. This process, which has been wellstudied in mice, is controlled by BMP and activin/nodal signaling[98,101,102]. Since bFGF in undifferentiated cells suppresses BMPsignaling [103], withdrawal of bFGF (Fig. 3 bottom, purple zone)potentiates the effects of the very low BMP levels in the media.Differentiation of the primitive streak population into definitiveendoderm is then induced by treatment of the cells with activin ornodal [104,105]. Activation of Smad2, in response to activin/nodalsignaling, results, through cooperation with other transcriptionfactors, in the expression of definitive endodermal genes, such asSox17, Cxcr4, Gata4, Gata6, FoxA2, Lhx1, and Gsc, and primitive streakgenes, such as Eomes, Mixl1, Fgf8, and Wnt3 [106–108]. When BMP isfurther added to ESCs under differentiation-inducing conditions inthe absence of serum, i.e. free from pluripotency growth factors suchas activin and bFGF, the Brachyury-expressing mesendoderm cells arefurther specified toward a committed mesodermal state, character-ized by expression of Flk1, the receptor for vascular endothelialgrowth factor (VEGF) [109]. In contrast, differentiation along theectodermal lineage is induced in the absence of BMP and TGF-βsignals. By adding inhibitors for BMP and TGF-β signaling, efficientinduction of neuroectodermal differentiation can be achieved[110,111].

ES

mEpiS

Blastocyst stage

Epiblast stage

Mesoderm Ectoderm

ES

TGF-β/Activin

TGF-β/Activin

BMP

BMP BMP

EndodermEndoderm Mesoderm Ectoderm

Activin

iPS

iPS

bFGF-dependent(primed state)

Differentiation

TGF-β inhibitorBMP inhibitor

TGF-β inhibitor

TGF-βinhibitor

Rep

rogr

amm

ing

TGF-β/Activin

Nodal

Mouse Human

ActivinNodal

BMP

BMP

ES iPS

bFGF-dependent(primed state)

LIF-dependent(naive state)

TGF-β inhibitor

TGF-β inhibitorBMP inhibitor

LIF

bFGF

Diff.factor

BMP

Nodal

BMP

Nodal

LIF-dependent(naive state)

Growth factorrequirements

Fig. 3. Roles of TGF-β family members in maintenance and differentiation of ESCs, and somatic cell reprogramming. ESCs derived frommouse blastocysts or human blastocysts havedistinct mechanisms to maintain pluripotency. Mouse ESCs (yellow circle) rely on LIF signaling, while human ESCs (green hexagon) require bFGF signaling. Fully reprogrammedmouse and human iPS cells (yellow and green squares) require the same signaling mediators as ESCs from the same species. Low BMP levels promote pluripotency of mouse ESCsand iPS (yellow) cells in the presence of LIF (orange zone), whereas TGF-β/activin signaling induces differentiation in the absence of LIF. Pluripotent mouse EpiSCs, derived fromepiblast stage embryos, require bFGF signaling, similarly to human ESCs and iPS (green) cells. These three cell types (green) require TGF-β/activin signaling together with bFGFsignaling (beige zone), while BMP signaling impairs their pluripotency. Activation of TGF-β/activin signaling in mouse ESCs causes transition from the naïve ESC to the primed EpiSCstate, when LIF signaling is blocked. Primed human ESCs (green hexagon) also revert to the naïve state (yellow hexagon) upon blocking TGF-β/activin signaling (dotted brownarrow). Differentiation of pluripotent cells into the three germ layers is induced by withdrawal of pluripotency factors such as LIF or bFGF (purple zone). Activin/nodal acts as anendoderm-inducing signal in the absence of bFGF. BMP and Nodal promote mesodemal differentiation, while blocking both activin/nodal and BMP signaling promotesneuroectoderm differentiation. BMP enhances somatic cell reprogramming (blue arrows), by inducing mesenchymal–epithelial transition. TGF-β type I receptor kinase inhibitorsalso promote iPS cell generation.


3.3. Reprogramming of somatic cells into iPS cells

Somatic cells can be reprogrammed into pluripotent, ESC-like cellsby overexpressing defined transcription factors. This iPS cell tech-nique was pioneered in mouse fibroblasts through ectopic expressionof four transcription factors, Oct4, Sox2, Klf4 and c-Myc [15].Depending on the reprogramming conditions, mouse somatic cellscan be reprogrammed into iPS cells with either mouse ESC or EpiSC(i.e. similar to human ESC) states, i.e. naïve and primed pluripotency,respectively (Fig. 3). Naive pluripotency is maintained by LIF and lowconcentration of BMP, while primed pluripotency is maintained bybFGF and TGF-β/activin signaling [16,112] (Fig. 3).

Interestingly, inhibition of the TGF-β/activin/nodal type I receptorkinases enhances iPSC induction and can eliminate the requirementto introduce Sox2 expression. Furthermore, a 24–48 h treatment ofintermediate, i.e. partially reprogrammed, iPS cells with this TGF-βreceptor inhibitor surprisingly induces Nanog expression to promotefull reprogramming [113,114]. Conversely, treating reprogrammingiPS cells with TGF-β, or introducing an activated type I TGF-β receptorlowers the efficiency of iPS reprogramming [114,115]. Furthermore,expression of miRNAs that inhibits expression of the TGF-β type IIreceptor and blocks TGF-β-induced EMT through downstream genes,enhances iPS cell reprogramming processes [115–117]. Therefore,

although TGF-β signaling is important for ES cell self-renewal, itinhibits reprogramming into iPS cells. Conversely, BMP signaling,which has been shown to induce mesenchymal-epithelial transi-tion, and thus opposes TGF-β stimulation in some contexts, appearsto promote reprogramming into iPS cells [118,119] (Fig. 3 bluearrow).

Regulation of TGF-β family signaling bymiRNAs is also seen in otherbiological processes besides iPS reprogramming. In human undiffer-entiated ESCs, miR-302 is one of the most abundant microRNAs. Ittargets many genes such as cyclin D1 and NR2F2 [120,121], butparticularly in the context of TGF-β family signaling, it targets theexpression of TGF-β signaling components including Lefty, as well asBMP signaling inhibitors [122,123]. In mouse liver stem cells, miR-23band miR-24-1 inhibit TGF-β signaling by downregulating Smad3expression, thus contributing to stemness [124]. In human hematopoi-etic progenitor cells, miR-24-1 inhibits TGF-β/activin signaling bytargeting the type I receptor ALK‐4 [125]. miR-21 induces adipogenicdifferentiation of human adipose tissue-derived mesenchymal stemcells by downregulating the expression of the TGF-β type II receptor[126]. All these examples emphasize that miRNAs play a determiningfactor in stem cell behavior by modulating the activation of thesignaling pathways. We will further discuss the role of TGF-β familysignaling in somatic stem cells in the next section.


4. Somatic stem cells

As with ESCs, the stemness of somatic or tissue stem cells isregulated by intrinsic properties of the stem cells, and non-autonomoussignals that are provided by their niche. Multiple external cues fromsoluble factors, membrane-bound molecules and extracellular matrixproteins influence the behavior, and ultimately direct the fatedetermination, of the somatic stem cells. To determine the keycomponents that regulate maintenance and differentiation of somaticstem cells, the signals released in the niche need to be elucidated. TGF-βsignaling by the niche cells has been implicated in themaintenance anddifferentiation of various types of somatic stem cells. In this section, wefocus on their roles inmammary, intestinal, skin,mesenchymal, skeletalmuscle, neural and hematopoietic stem cells.

4.1. Mammary stem cells

Mammary stem cells are the source of epithelial cells in thegrowth and development of the mammary gland during puberty andgestation. They have been identified in, and isolated from human andmouse mammary tissue, and from cell lines derived from themammary gland [127–129]. Single mammary stem cells can giverise to both luminal and myoepithelial cells of the gland, and wereshown to regenerate entire mammary gland structures in mice [130].Whereas many signaling pathways, such as Hedgehog, EGF, Wnt, FGF,and Notch signaling, control mammary stem cell proliferation andsurvival, TGF-β signaling appears to play a critical role in stemness ofnormal mammary epithelial cells [131,132].

Increasing evidence links stemness ofmammary epithelial cellswithEMT. EMT is a key event in various differentiation contexts in normalembryogenesis, cancer invasion and metastasis, and fibrosis [133,134].EMT occurs when epithelial cells lose their epithelial characteristics,and acquire mesenchymal phenotypes, resulting in increased motility.TGF-β is a potent inducer of EMT in normal mammary epithelial cells[8,135,136], and many studies have established crucial roles for TGF-β-induced EMT in tumor progression [133,137–140]. EMT in immortal-ized human mammary epithelial cells results in acquisition ofmesenchymal markers, such as increased fibronectin and vimentinexpression, increased invasiveness, and a CD44high/CD24low antigenicphenotype, which also characterizes mammary CSCs. Cells that haveundergone EMT gain a higher potential to form mammospheres, aproperty associated with mammary epithelial stem cells [30]. Further-more, TGF-β and Wnt signaling cooperate to induce activation of theEMT program, and function in an autocrine fashion to maintain theresulting stem cell state [31]. These findings demonstrate thatdifferentiated epithelial cells have a potential to dedifferentiate andgain stem cell properties through EMT.

4.2. Intestinal stem cells

Intestinal stem cells divide throughout life and give rise to theepithelial cells lining the surface of the small and large intestines [141].Intestinal stem cells reside near the base of the stem cell niche, thecrypts of Lieberkühn. Genetic, inducible fate mapping studies haveidentified two principal epithelial stem cell pools. One pool consists ofcolumnar Lgr5-expressing cells that proliferate rapidly and residepredominantly at the base of the crypts [142]. The other pool consists ofBmi1-expressing cells that proliferate slowly and largely reside abovethe crypt base [143]. Both stem cell populationswere shown to give riseto differentiated epithelial cells [144,145], although, the functionalrelevance of having two different types of stem cell populations, andtheir relative contributions remain to be fully evaluated.

The proliferation and differentiation of the intestinal stem cells areregulated by factors secreted from the underlying mesenchymallayer, which include fibroblasts, enteric neurons, blood vessels, andextracellular matrix. Wnt signaling stimulates the proliferation of the

stem cells and the transiently amplifying cells in the crypts [146],while BMP signaling acts as a negative regulator of stem cellproliferation in the crypts. BMP4 is highly expressed in the intravillusmesenchyme [147], and phosphorylation and nuclear localization ofthe BMP-specific effector Smads have been observed in differentiatedvillus epithelial cells and intestinal stem cells [148]. These observa-tions suggest that BMP signaling from the mesenchyme acts onadjacent, non-proliferating epithelial cells and slowly proliferatingstem cells. Mutations in the genes encoding Smad4 [149] or BMPRIA/ALK-3 [150] have been linked to juvenile polyposis in humans.Furthermore, exogenous expression of the BMP antagonist noggin inthe mouse intestine leads to de novo ectopic formation of normalcrypt-villus units, and, at later stages, formation of a complexarchitecture of branching villi with dilated cysts, similar to theintestinal phenotype of human juvenile polyposis [147]. In addition,conditional deletion of the Bmpr1a gene in crypts in mice disturbs theregeneration of intestinal epithelial cells, with expansion of stem andtransiently amplifying cells, eventually leading to a type of intestinalpolyposis, reminiscent of human juvenile polyposis [148]. These datasuggest that BMP signaling is important for a balanced control of stemcell self-renewal, specifically in maintaining quiescence of Bmi1-positive slowly proliferating stem cells, and in terminal differentia-tion of intestinal epithelial villi, possibly by opposing Wnt signaling,which stimulates proliferation of the Lgr5-positive stem cellpopulation.

4.3. Skin stem cells

The skin epidermis and its appendages provide a protectivebarrier to harmful microbes and prevent dehydration. To performtheir functions and maintain tissue homeostasis, skin continuouslyrejuvenates by replenishing old cells with new cells that differentiatefrom skin stem cells. Based on their localization and function, threedifferent types of skin stem cells have been discerned, i.e. epidermal,hair follicle, and melanocyte stem cells. We focus on hair follicle stemcells (HFSCs) and melanocyte stem cells, since recent studieshighlighted the role of TGF-β signaling in the maintenance of theirstemness.

4.3.1. Hair follicle stem cellsThroughout adult life, HFSCs undergo dynamic, synchronized

cycles of degeneration (catagen), quiescence (telogen), and regener-ation (anagen) [151–153]. In the telogen phase, HFSCs are quiescentand reside in a specialized microenvironment, called the hair bulge[154]. In telogen, the base of the bulge is directly adjacent to theunderlying dermal papilla, a signaling center for HFSCs.

TGF-βs are expressed by hair follicles in catagen, and their role inthis state, through induction of apoptosis, has been well established[155,156]. However, targeted inactivation of the individual genesencoding the three TGF-β ligands resulted in differential effects onembryonic hair follicle development. For example, mice with adisruption of the Tgfb1 gene show slightly advanced hair follicleformation, while lack of the Tgfb3 gene does not have any effects.Tgfb2−/− mice exhibit a profound delay of hair follicle morphogen-esis, with a 50% reduced number of hair follicles [157]. Thus, the rolesof TGF-β ligands in HFSCs appear to be very diverse. TGF-β2expression in the dermal papilla acts as a critical signal in promotingHFSC regeneration, and HFSCs that cannot sense TGF-β exhibitsignificant delays in hair follicle regeneration. Additionally, TGF-β2antagonizes BMP signaling in HFSCs by enhancing the expression ofTmeff1, an antagonist of the BMP pathway, thus restricting andlowering BMP responsiveness in the niche [28].

While TGF-β signaling needs to be activated for the transition ofthe dermal papilla from the quiescent to the regeneration state, BMPsignaling supports the maintenance of quiescent HFSCs, reminiscentof a similar role in intestinal stem cells. When the Bmpr1a gene is


inactivated in postnatal skin epithelium, the quiescent HFSCs areactivated to proliferate, leading to loss of slowly proliferating cells[158]. In line with this observation, BMP inhibitory factors from thesurrounding niche cells have been implicated in the induction ofanagen [159–161]. Conversely, expression of a constitutively activeBMPRIA in skin promotes premature hair follicle differentiation,further supporting the notion that BMP signals are critical inmaintaining the properties of quiescent stem cells [158].

4.3.2. Melanocyte stem cellsMelanocytes are the pigment-producing cells in the skin, hair and

eye that determine their color. In the hair bulge, melanocyte stemcells are found where the HFSCs reside [162]. Their differentiationstate is defined by the expression of genes involved in pigmentation,which is driven by the master transcription factor Mitf [163,164].Melanocyte stem cells are usually in a quiescent, non-cycling stateand are activated only at early anagen to undergo stem cell division. Aprecise control mechanism through Mitf and Bcl2 is important in themaintenance of stemness of melanocytes, and to avoid depletion ofthe stem cell source, which leads to hair graying [165].

Expression of TGF-β ligands by niche cells in the hair bulge [166]plays an important role in the physiology of the melanocyte stem cells,as they induce cell cycle arrest in melanocyte stem cells to let thementer dormancy [29]. Furthermore, TGF-β induces downregulation ofMitf expression, consequently preventing expression of melanogenicdifferentiation genes [29]. As a result, mice that specifically lackexpression of the TGF-β type II receptor in melanocytes, exhibit mildbut accelerated hair graying [29].

Interestingly, the HFSCs serve as melanocyte stem cell niche byexpressing TGF-β [167]. Mice lacking collagen17α1, which exhibithair loss due to the lack of the HFSC maintenance program, retainTGF-β1/2 expression in the bulge area of hair follicles at 5 weeks, butgradually loose TGF-β expression by 8 weeks. Correlating withdefects in HFSC maintenance upon loss of TGF-β signaling, aberrantmelanocyte differentiation starts at 12 weeks in col17a1−/− mice[167], suggesting that melanocyte stem cells are maintained by beingadjacent to normal HFSC, which produces TGF-β as a niche signal formelanocyte stem cells. In this context, TGF-β stimulates hairregeneration by activating the cell cycle of HFSCs, while also inducingthe entry to the quiescence of melanocyte stem cells to potentiatetheir function as a reservoir of a pigment-producing organ for longerperiods.

4.4. Mesenchymal stem cells

Mesenchymal stem cells (MSCs) have been isolated from variousadult tissues including connective tissue, adipose tissue, muscle, bonemarrow and teeth, blood, placenta and umbilical cord [168,169]. Thedifferentiation potential of MSCs depends on the local tissue environ-ment from which they are isolated, and, in vivo, is dictated by theirlocalization and microenvironment. For example, bone marrow-derivedMSCs are capable of differentiating into osteoblasts, adipocytes,chondrocytes, muscle cells, and hematopoiesis-supporting stromalcells, but not into hematopoietic cells [170]. Muscle-derived MSCs areable to give rise tomyogenic, osteogenic, chondrogenic, and adipogeniccells [171]. Human MSCs derived from bone marrow can be expandedmore than a billion-fold in culture without losing their stem cellcapacity. Owing to their potential for clinical applications, MSCs haveattracted much attention [172].

Proliferation of human MSCs is stimulated by Wnt or TGF-βsignaling [173–175]. TGF-β1 induces Smad3-dependent nuclearaccumulation of β-catenin in MSCs, which is required for stimulationof MSC proliferation. On the other hand, BMP2 antagonizes Wnt3asignaling and inhibits proliferation of bone marrow-derived MSCsthrough interaction of BMP-activated Smad1/5 with Dishevelled-1, acomponent of the Wnt signaling pathway [176].

TGF-β signaling has also been implicated in directing the differen-tiation fate of MSCs [177]. BMPs induce differentiation of mesenchymalcells into chondroblasts or osteoblasts in vitro. TGF-β and activin alsoprovide competence for chondroblast differentiation at early stages,while TGF-β inhibits osteoblast maturation at late stages in differenti-ation [177]. Consequently, pharmacological inhibition of TGF-β/activinsignaling strongly enhances osteoblast maturation [178]. These inhib-itory effects are mediated by induction of expression of inhibitorySmads, such as Smad6, which in turn represses BMP signaling [179]. Inaddition, BMP7 was shown to induce the generation of brown fat fromMSCs in the absence of the normally required hormonal induction[180]. Since brown fat is specialized in energy expenditure and cancounteract obesity, a therapeutic strategy based on the use of BMP7may be suggested as a treatment of obesity.

While MSCs can differentiate into skeletal, smooth and cardiacmuscle cells [181,182], many studies have focused on their potential togive rise to cardiomyocytes for cardiac tissue regeneration afterinfarction [183–189]. A correlation between in vivo transplantation ofMSCs and repair of scarred myocardium after myocardial infarction inrats [190] has led to the idea that the implantation of MSCs intomyocardium could be a useful clinical approach. Treatment of humanMSCs with TGF-β induces the expression of the Notch ligand Jagged 1along with cardiomyocyte markers, including α-smooth muscle actin,calponin 1, and myocardin. Increased expression of these genesdepends on activation of Smad3 and Rho kinase. In addition, preventingJagged 1 expression blocks the expression of cardiomyocyte genesshown above, suggesting that Jagged 1 plays an important role in TGF-β-induced expression of cardiomyocyte marker genes [191]. Thesestudies implicate a role of the microenvironment in the induction ofdifferentiation of MSCs along the cardiomyocyte lineage, and TGF-βsignaling is a candidate key niche factor to promote this differentiation.Further studies in animal models will allow us to understand theunderlying mechanisms of how in vivo niche factors, such as TGF-β,control MSC differentiation.

4.5. Skeletal muscle stem cells

Muscle stem cells show myogenic potential in response to environ-mental cues, and have been isolated from skeletal muscle [192–194].They can be categorized into satellite cells, mesoangioblasts (vessel-associated stem cells), side population cells, muscle-derived stem cells,pericytes, and CD133-positive stem cells [194–199]. Among these stemcell types, satellite cells have been extensively studied. Since a fairamount of knowledge has accumulated in the past few decades, the roleof TGF-β signaling in satellite cells has been well documented.

Satellite cells are located between the basal lamina and thesarcolemma of the muscle fiber, and become activated for prolifer-ation and differentiation in response to trauma, mechanical injury orbiological stimuli. Satellite cells are a heterogeneous cell populationcomprised of stem cells and committed progenitors, based onexpression of Pax7 or Myf5 [200]. Once stimulated, satellite cellsproliferate, differentiate into myoblasts, and fuse with existingmuscle fibers or, alternatively, create fibers de novo [193,201].

In skeletal muscle, TGF-β family members have inhibitory effectson both muscle development and postnatal regeneration of skeletalmuscle [202]. TGF-β1 represses skeletal muscle-specific gene expres-sion, and regulates proliferation in satellite cells [203–208]. Myostatinis another important TGF-β family member in skeletal muscledevelopment. Mice lacking myostatin expression exhibit increasedmuscle mass and fiber hypertrophy, which are indicative of theinhibitory effects of myostatin on muscle development [209–211].Interestingly, short-term inhibition of myostatin in aged mice thatnormally retain less muscle regeneration potential, enhances muscleregeneration and activates satellite cell proliferation [212]. In agedhuman skeletal muscle, satellite cells are maintained but fail to beactivated upon environmental stimuli. These cells exhibit a decline in


Notch activation due to decreased Erk MAPK signaling [213].Moreover, aged human muscle produces excessive TGF-β, but notmyostatin, which induces unusually high levels of Smad3 activationin resident satellite cells and interferes with their regenerationcapacity [213].

Increased TGF-β signaling is also apparent in skeletal muscle offibrillin-1-deficient and dystrophin-deficient mice, models of humanMarfan syndrome and Duchenne muscular dystrophy, respectively.Antagonizing TGF-β signaling rescues skeletal muscle regeneration infibrillin-1-deficient mice and restores the regeneration program inthese mice [214]. Furthermore, fibrillin-1-deficient mice exhibitincreased satellite cell numbers when TGF-β is antagonized [215].These findings suggest that TGF-β and myostatin inhibit satellite cellactivation, and thus lead to a failure of skeletal muscle regeneration.

Signaling by TGF-β or myostatin converges in activation of Smad2and Smad3. Several studies have elucidated roles of Smad2 andSmad3 in TGF-β1-mediated myogenic inhibition and in normaldifferentiation [47,216]. Smad3, but not Smad2, is the key mediatorof TGF-β inhibition of myogenesis, through inhibition of MyoD andother myogenic bHLH transcription factors that drive myogenicdifferentiation [47]. Smad3 associates physically with MyoD, inhibit-ing the transactivation properties of MyoD [47], and withMEF2C, thusdecreasing its transcription activity [217]. MEF2C plays a role in thelate stages of myogenic differentiation, and conditional deletion inskeletal muscle affects sarcomere integrity [218].

To fully understand the molecular mechanisms of the TGF-β andmyostatin signaling pathways in skeletal muscle stem cells, furtherwork in different stem cell types will be needed. Profiling TGF-β ormyostatin target genes at a genome-wide level may give furtherinsights into common and divergent roles of these differentiationfactors.

4.6. Neural stem cells

The self-renewing neural stem cells give rise to neural progenitorcells and eventually to neurons, astrocytes and oligodendrocytes[219]. Adult neurogenesis, resulting in the formation of new neurons,has been demonstrated in brains of adult mice, songbirds andprimates, including humans, specifically in the subventricular zone,which lines the lateral ventricles, and the dentate gyrus of thehippocampus [220]. These observations led to the identification ofneural stem cells in these structures, although the presence of trueself-renewing stem cells there has been debated [221]. Under somecircumstances, such as following tissue damage in ischemia, neuro-genesis can be induced in other brain regions as well, including in theneocortex. Neural stem cells can be propagated in vitro as neuro-spheres, and differentiated into both neuronal and glia cells [219];however, some studies suggest that this differentiation behavior isinduced by culture conditions [222,223]. A better understanding ofthe growth factors or stimuli that control the proliferation and lineagedecisions of these cells is warranted.

TGF-β signals play important roles in the maintenance and growthof neural stem cells [224]. Targeted inactivation of TGF-β type IIreceptor gene in the mid/hind brain enhanced the self-renewal, butnot the multipotency, of mouse neural stem cells, resulting in anenlarged midbrain. Ectopic expression of FGF and Wnt ligands, suchas FGF8 and Wnt1, was observed in these mutant brains, suggestingthat TGF-β signaling antagonizes FGF and Wnt signaling, thuscontrolling the size of the midbrain by negatively regulating self-renewal of neural stem cells.

BMPs induce a variety of biological responses in embryonic neuralstem cells [225]. BMP2/4 acts as a neuroepithelial proliferation signalat very early stages of embryonic central nervous system develop-ment [226]. Later in development, BMPs induce neuronal andastroglial differentiation of neural stem cells [227,228]. BMP signalshave also been implicated in the maintenance of neural stem cells in

adult brain. In the subventricular zone, where somatic neural stemcells reside [229], multiple cell types can be discerned, such asmigrating neuroblasts, immature precursors, astrocytes, and ependy-mal cells. These astrocytes were shown to act as neural stem cells inboth normal and regenerating brain [230]. Among the many signalingmolecules that are expressed in the subventricular zone, BMPs inducethe generation of astrocytes at the expense of new oligodendrocyteand neuron generation. In contrast, noggin is highly expressed inependymal cells, which do not act as neural stem cells [231]. Localnoggin expression may contribute to the formation of a neurogenicniche for stem cells in the subventricular zone, as it promotesdifferentiation. These findings suggest that BMPs play a role in themaintenance of neural stem cells, i.e. astrocytes, by inhibiting theircommitment to neurons and oligodendrocytes.

4.7. Hematopoietic stem cells

Hematopoietic stem cells (HSCs) are found in bone marrow, andgive rise to all blood cell types. Although various methods are used topurify HSCs [232], these cells are usually not defined by phenotype,but rather by their ability to reconstitute a hematopoietic cellpopulation in bone marrow [233,234]. Accordingly, HSCs have beenclassified by their lineage differentiation capacity, assessed by serialtransplantation of single HSCs [235]. Another classification of HSCs isbased on repopulation kinetics in mice transplanted with limitingdilutions of whole bone marrow [236–238]. These findings indicatethat each subtype of HSCs displays different responsiveness to themicroenvironment, and thus different responses to growth factors intheir niche.

TGF-β signaling has been implicated in the regulation ofquiescence of HSCs [239]. Since TGF-β is produced in a latent formby a variety of cells, and multiple mechanisms can lead to TGF-βactivation, the mechanism of activation of latent TGF-β in the contextof HSC quiescence remains to be resolved. Non-myelinating Schwanncells were shown to be in contact with a substantial proportion ofHSCs in bone marrow, and to mediate activation of TGF-β. Thesefindings suggest that glial cells maintain HSC quiescence by limitingactivation of latent TGF-β as components of a bone marrow niche[240].

In cell culture, TGF-β signaling-deficient HSCs have a higherproliferative capacity, whereas the quiescence and maintenance ofHSCs in vivo may depend on TGF-β signaling [241,242]. Furthermore,the response of HSCs to TGF-β stimulation is biphasic; high concentra-tions of TGF-β inhibit, while low concentrations stimulate HSCproliferation [243]. Additionally, TGF-β signaling results in differenteffects on myeloid-biased (My-) and lymphoid-biased (Ly-) HSCsubtypes in various assays, such as colony formation from single cells,cell cycle analyses, and proliferation in vivo [244]. These observationssuggest a mechanism for differential regulation of distinct HSCsubtypes. Differential responses between subtypes of leukemic cells,such as leukemic stem cells and non-stem cells, toward environmentalcues can also be explained by cell heterogeneity.

5. Cancer stem cells

Cancer stem cells (CSCs) are a small population of self-renewing cellswith the ability to initiate tumor formation and give rise to heteroge-nous cancer cells. Increasing evidence highlights the importance of CSCsin cancer progression [23,245]. CSCs are thought to confer resistance tochemotherapy or radiotherapy, leading to recurrence and metastases[24]. Thus, characterizing the properties of CSCs and themechanisms oftheir regulation by signaling pathways is of high importance. CSCs arethought to be maintained by a cancer cell microenvironment that isconceptually similar to a normal stem cell niche [26]. Growth factorsand cytokines released from the cancer microenvironment are thoughtto regulate CSC maintenance, differentiation and function. Recent


studies have revealed critical roles of TGF-β family signaling in CSCmaintenance and differentiation in several types of tumors (Table 1).Their roles are strikingly similar to those in normal somatic stem cellmaintenance and differentiation.

5.1. Epithelial cancer stem cells

Epithelial cancers, also called carcinomas, are cancers formed fromepithelial tissues such as skin, breast, gastrointestinal tract, kidneyand bladder. Epithelial cancers make up about 85% of all the cancertypes. Recent evidence suggests plasticity in epithelial cancer cellsthat allows non-CSCs to acquire CSC-like properties, and vice versa.Thus, the hierarchical, unidirectional model in which CSCs are oftendescribed as cells at the apex of a hierarchy is now being replacedwith a model in which signals from the cancer microenvironment cancreate new CSCs from non-CSCs [246] (Fig. 4A). Several reportsdemonstrate that TGF-β-induced EMT can lead to dedifferentiationand gain of stem cell-like properties in both cancer and normal cells[247] (Table 1, Fig. 4A).

5.1.1. Breast cancer stem cellsIn human breast cancers, a small subpopulation of cancer cells

with a CD44high/CD24low antigenic phenotype was shown to exhibitCSC properties [248]. High aldehyde dehydrogenase 1 (ALDH1)activity also identifies cell populations with CSC properties [249].Interestingly, stem cell-like cells isolated from mouse or humanmammary carcinomas, as well as normal mammary glands, wereshown to express EMT markers. Transformed human mammaryepithelial cells that have undergone EMT show an increased ability toformmammospheres, soft agar colonies and tumors, which correlateswith their CSC properties [30]. Additionally, in immortalized humanmammary epithelial cells, oncogenic Ras and TGF-β1 cooperate toproduce CD24low stem cell-like cells from CD24high differentiatedcells [250]. These observations link TGF-β-mediated EMT withacquisition of the CD24low CSC phenotype. Furthermore, a mesen-chymal cell subpopulation spontaneously arises from immortalizedhuman mammary epithelial cells through EMT. In these mesenchy-mal cells, autocrine TGF-β signaling is necessary for maintainingthe mesenchymal state and tumorigenicity [31]. In this cell system,

Table 1Reported cancer stem cells and corresponding literature.

Cancer type Roles of TGF-β family signaling References

Breast cancer TGF-β-induced EMT maintains stem-like properties [30,31]TGF-β reduces the size of the SP fraction and theability of tumor formation

[252]

BMP2/7 decrease ALDH+/CD44high/CD24low stemcells

[251]

Pancreascancer

Nodal/activin signaling drives self-renewal andtumorigenicity

[255]

Diffuse-typegastric cancer

TGF-β decreases the CSC fraction [269,270]

Colorectalcancer

BMP4 induces differentiation, apoptosis andchemosensitization of CSCs

[258]

Squamous cellcarcinoma

TGF-β/TβRII signaling restricts self-renewal andexpansion of CSCs

[268]

Ovarian cancer TGF-β-induced TG2 induces EMT and a CSCphenotype

[271]

Prostate cancer BMP7 induces senescence in CSCs by activating p38MAPK and increasing the expression of p21 andNDRG1

[272]

Glioma TGF-β maintains tumorigenicity and stemness [276,277]BMP4 inhibits proliferation and inducesdifferentiation of human glioblastoma stem cells

[274]

Leukemia TGF-β–FOXO pathway maintains stem cell-likeproperties

[282]

Smad4 negatively regulates leukemia initiation andmaintenance induced by HOXA9 gene or NUP98–HOXA9 fusion oncogene

[283]

in which BMP antagonizes TGF-β-induced EMT, BMP signaling isdecreased and TGF-β signaling is increased, when compared tothe normal epithelial cell population [31]. Supporting thesereports, BMP2/7 antagonizes TGF-β-mediated EMT, and decreasesthe frequency of ALDHhigh/CD44high/CD24low breast CSCs. Bonemetastasis, formed after intra-cardiac injection of MDA-MB-231breast cancer cells in Balb/c nu/nu mice, is also decreased withBMP2/7 treatment [251]. These reports shed light on the notion thatnormal mammary stem cells and breast CSCs may share similarmechanisms, which rely in part on TGF-β signaling-mediated EMT, toestablish and maintain stemness. The concept of epithelial cancerplasticity through EMT may provide explanations for how multi-stepcancer progression is achieved. If oncogenic changes that normallyoccur in non-CSCs can be introduced to CSCs by dedifferentiationthrough EMT, CSCs with second oncogenic mutations can drivemetastasis and resistance to chemotherapies and radiotherapies.Thus, TGF-β-induced EMT may enhance cancer progression byinducing not only an invasive phenotype, but also dedifferentiationof non-CSCs into CSCs.

In contrast to reports demonstrating that TGF-β signaling inducesexpansion of the CSC population and promotes cancer progression,TGF-β has also been shown to decrease the number of CSCs andinhibit tumor formation. In immortalized and transformed humanbreast epithelial cells derived from MCF10A cells, TGF-β reduces thesize of the side population cell fraction, which exhibits properties ofCSCs [252]. In this model, TGF-β promotes differentiation of CSCpopulation by decreasing the expression of Id1, which induces self-renewal and inhibits differentiation in many tissues. By enhancingdifferentiation and reducing the number of CSCs, TGF-β inhibitstumor formation [252]. These results suggest complex roles of TGF-βin the regulation of CSCs, and different effects of TGF-β on CSCpopulations from different tumors, that may result from differentialexpression of signaling regulators in different populations. Furtherstudies on the roles and mechanisms of TGF-β in the regulation ofCSCs using different tumor types are needed to develop treatmentsbased on activities of TGF-β signaling.

5.1.2. Pancreas cancer stem cellsPancreas cancers also harbor distinct subpopulations of CSCs with

self-renewal capacity, abilities to differentiate and initiate tumori-genesis [253,254]. Human pancreas CSCs and stroma-derived pan-creatic stellate cells express nodal and activin, important regulators ofESC pluripotency, and activated nodal/activin signaling is required forself-renewal and tumorigenicity of pancreas CSCs [255]. Pharmaco-logical inhibition of the nodal/activin type I receptor kinases reducesself-renewal and tumorigenicity of these CSCs, and enhances theirsensitivity to chemotherapy [255]. These results suggest that, tomaintain their self-renewal capacity, tumorigenicity and resistance tochemotherapy, pancreas CSCs may use the same signaling pathwaysas those that are involved in human ESC pluripotency. Inhibition ofnodal/activin signaling, combined with chemotherapy, may provide atherapeutic strategy for targeting pancreas CSCs. On the other hand,about 50% of patients with pancreas cancer bear impaired Smad4function, which suggests other key mechanisms besides nodal/activinsignaling to maintain pancreas CSCs. Further studies will be needed todefine several potential therapeutic strategies that will be chosenaccording to the mutations.

5.1.3. Colon cancer stem cellsIn human colon cancer, CSCs were initially identified as a CD133+

cell subpopulation [256,257]. These CSCs can be expanded as tumorspheres using serum-free media, but, when grown in spheres inthe presence of serum, they acquire epithelial markers and losetumorigenicity.

BMPs promote differentiation of normal colon stem cells, andinactivation of BMP signaling confers increased tumorigenesis.

HSCLSC

Normal HSC niche

homing to alternative niche

osteoblast

endothelial cells

osteoclast

Niche supports LSCs

oncogenic change

CSC

Differentiation

Self-Renewal

non-CSC

EMT

oncogenic change

A

non-CSC with a new mutation

CSC

CSC witha new mutation

B

EMT

LSC niche

Proliferation

Formation of tumorat distant site

non-CSC

Proliferation

LeukemiaCarcinoma

Fig. 4. Cancer stem cells (CSCs) in tumor initiation and progression. (A) Plasticity of epithelial cancer cells through epithelial–mesenchymal transition (EMT) and its role in tumorprogression. Epithelial CSCs have the ability to self-renew (orange arrow) or to differentiate (blue arrow) into different types of non-CSCs (blue, green), and acquire oncogenicchanges. Non-CSCs can undergo EMT (red arrow), thus acquiring a mesenchymal phenotype and stem cell-like properties, and generating new CSCs from non-CSCs. A cancer cell innon-CSC population may acquire an additional oncogenic change (pink arrow) that promotes tumor progression. Dedifferentiation of this non-CSC (pink) through EMT generates aCSC with a second oncogenic mutation (pink), which has a more malignant phenotype and drives expansion, invasion and metastasis of the tumor. (B) Leukemic stem cells (LSCs)and their niche. The interactions between LSCs and their niche have much in common with the interactions between normal hematopoietic stem cells (HSCs) and their bonemarrow niche. The normal niche contains HSCs and supporting cells, such as osteoblasts. Oncogenic changes (pink arrow) in HSCs can lead to the generation of LSCs. LSCs alter theproperties of normal niche cells, and the niche adapts to LSCs. Growth factors and cytokines released from the altered niche support the LSCs to maintain their stemness. Signalsfrom the niche also induce additional genetic or epigenetic changes in LSCs. LSCs that are no longer dependent on the original niche survive in the blood vessel and may home to analternative niche, leading to expansion of leukemic cells.


Inactivation of BMP signaling by impaired BMPRIA or Smad4 function isoften observed in human colon cancers [147,149,150]. Consistent withthese observations, BMP4, which is expressed in non-CSCs, but not inCSCs, induces differentiation, apoptosis and sensitivity to chemotherapyin the CSC population of human colorectal cancer cells that do not haveimpaired Smad4 function and constitutive activation of PI3K [258]. Thiseffect of BMP4 ismediated through inhibition of the PI3K–Akt pathway,which is normally activated in CSCs, through Smad-dependent andSmad-independent mechanisms. These and other findings suggest thatthe same signaling pathways can enhance differentiation of bothnormal and cancer stem cells, and that BMP4 may be used as atherapeutic agent against CSCs in colon cancers.

5.1.4. Melanoma stem cellsSimilarly to other types of CSCs, malignant melanoma stem cells

have the ability to initiate new tumors and generate a heterogeneouspopulation of cells that includes CSCs [259]. Cell-surface markersspecific for melanoma stem cells, such as CD20 and CD133 have beenidentified [260,261], but the correlation between expression of thesemarkers and the cells' capacities for self-renewal and tumorigenesisremains to be elucidated. More recent reports have shown that a highproportion (at least 25%) of melanoma cells isolated from patients wereable to initiate tumorigenesis from a single cell [262]. Furthermore,none of the 22 heterogenously expressed markers that are commonlyused to identify CSCs, can isolate tumorigenicmelanoma stem cells, andany large subpopulation ofmelanoma cells does not lack tumorigenicity[262]. Moreover, reversible epigenetic regulatory mechanisms allowmelanoma cells to readily convert between a CSC state, which isquiescent and resistant to therapy, and a non-CSC state that cannot

initiate tumor formation [263]. This phenotypic plasticity in melanomacells may represent an extreme example of functional plasticity ofepithelial cancers [264]. Signals from melanoma microenvironmentsplay a key role in regulating this fate change from a stem cell-like to anon-stem cell-like state. Nodal, a regulator of human ESC pluripotency,is expressed in human melanoma cells, and is required, at least in part,for tumor cell plasticity and aggressiveness [265–267]. Anti-nodalantibodies can inhibit nodal-induced plasticity in vitro and lungcolonization in mice [265,267]. Nodal expression is not detected innormal melanocytes, suggesting deregulated nodal signaling in mela-noma cells. Nodal may represent a novel prognostic and therapeutictarget for human melanoma.

5.1.5. Other epithelial cancersTGF-β family signaling also plays important roles in other types of

epithelial CSCs. In squamous cell carcinoma, a subpopulation ofα6β1hiCD34hi cells is thought to function as CSCs. TGF-β signalinginhibits self-renewal and expansion of α6β1hiCD34hi cells in aTgfbr2−/− mouse model of squamous cell carcinoma [268]. In humandiffuse-type gastric carcinoma, TGF-β from the tumor microenviron-ment decreases the side population or ALDH1+ cells, which show CSCproperties, and inhibits tumor formation in immunodeficient mice[269,270]. In primary human ovarian cancer cells, TGF-β secreted inthe tumor microenvironment induces expression of tissue transglu-taminase 2 (TG2). TG2 induces EMT and spheroid formation, whichcorrelate with the CSC properties, to enhance metastasis [271]. Inhuman prostate cancer, BMP7, secreted from bone stromal cells,induces reversible senescence of CSCs. BMP7 signaling induces theCSCs to become dormant, in part by activating p38 MAPK and


increasing the expression of the cell cycle inhibitor p21CIP1 and themetastasis suppressor NDRG1 (N-myc downstream-regulated gene 1)[272].

5.2. Glioma stem cells

Human gliomas, which are derived from multipotent neural stemcells, neural progenitor cells or glial progenitor cells [25], contain asmall fraction of glioblastoma stem cells (GSCs), which retain manysimilarities to neural stem cells, including the abilities to self-renewand differentiate into neural and glial cells. GSCs express specificmarkers such as CD133 and nestin that are commonly expressed innormal neural stem cells [273]. Similarly to BMP-induced astrocyticdifferentiation of embryonic neural stem cells, BMP4 inhibits theproliferation of GSCs, and induces their differentiation into gliomacells with an astroglial-like fate [274]. Consequently, BMP4 deliveredto the tumor in vivo, blocks tumor growth in mice followingintracerebral grafting of human glioblastoma cells [274]. However,BMP-induced astroglial differentiation appears to be impaired in asubpopulation of GSCs, due to epigenetic silencing of the BMPR1Bgene [275]. These results suggest that new oncogenic alterations thatcould be beneficial for the expansion of GSCs may selectivelyaccumulate under some conditions. These deregulated signals inGSCs may reveal themselves as possible targets for therapeuticapproaches.

Increased TGF-β1 expression is commonly seen in malignantglioma cells, similarly to upregulated TGF-β1 expression in carcino-mas and other tumors of neuro-ectodermal origin. Recent reportshave revealed key roles of TGF-β signaling in the maintenance of stemcell-like properties and the tumor initiating ability of GSCs. TGF-βmaintains the tumorigenicity of human GSCs through Smad-mediated induction of LIF expression, which then, in an autocrineway, activates the JAK–STAT signaling pathway in GSCs, to increaseself-renewal and decrease differentiation [276]. TGF-β signalingmaintains stem cell properties of human GSCs through Smad-dependent induction of SOX4 expression and subsequent inductionof SOX2 expression. SOX2 is an ESC self-renewal gene, which regulatesthe expression of other transcription factors that are critical formaintaining stem cell-like properties of ESCs and neural stem cells[277]. These findings show that TGF-β signaling maintains GSCs'population and promotes glioblastoma formation. Accordingly,pharmacological inhibition of the TβRI kinase can augment theresponse to radiation treatment in a mouse model of orthotopic,intracranially implanted human glioblastoma cells [278]. By coordi-nately increasing apoptosis of glioma cells and GSCs, while blockingDNA damage repair, invasion, mesenchymal transition, and angio-genesis, the TβRI kinase inhibitor blocks glioma expansion [278]. Thecombination of radiotherapy with inhibition of TGF-β signaling maybe a therapeutic strategy for gliomas.

5.3. Leukemic stem cells

Leukemic stem cells (LSCs) were first identified as a phenotypi-cally and functionally distinct cell population in acute myeloidleukemia (AML) [279]. Various studies examined the role of thestem cell niche in the control of LSCs. In both AML and chronicmyeloid leukemia (CML), the interaction of the stem cell markerCD44, which is expressed in LSCs, with hyaluronic acid moieties at thesurface of stromal cells in bone marrow, is essential for adherence ofLSCs to the niche [280,281]. This mechanism resembles that of theinteraction between normal HSCs and the vascular niche, againrevealing parallels between normal HSCs and LSCs (Fig. 4B).

TGF-β signaling plays crucial roles in the maintenance of LSCs. Inhuman CML, TGF-β produced by LSCs, or in the tumor microenviron-ment, is thought to be required for survival, proliferation, andmaintenance of stem cell-like properties of LSCs, through activating

Akt signaling, which induces nuclear localization of FOXO3a [282].Regulation of Akt activation and nuclear localization of FOXO3a inresponse to TGF-β has also been shown in normal HSCs [239],demonstrating a commonmechanism for normal HSCs and LSCs. TGF-β-induced FOXO3a nuclear localization was specific to LSC, and wasnot observed in the non-LSC cell population [282], suggesting thatTGF-β has different effects on FOXO3a activation in LSCs and non-LSCs. Thus, inhibition of TGF-β-FOXO signaling may provide apossible therapeutic approach for the treatment of human CML.

In contrast to the stimulatory effect of TGF-β signaling on CML stemcell maintenance and proliferation, TGF-β-induced Smad signalingimpairs initiation and progression of leukemia in amousemodel of AML[283]. In amousemodel of human AML induced byHOXA9 or the fusiononcogene NUP98–HOXA9, Smad4 binds and stabilizes Hoxa9 in thecytoplasm, thus inhibiting nuclear localization of Hoxa9. Since nuclearHoxa9 induces transformation of normal hematopoietic stem orprogenitor cells to LSCs, Smad4 can negatively regulate the transfor-mation of HSC into LSC [283]. Loss of Smad4 increases Hoxa9 levels inthe nucleus, enhancing the transformation of HSC into LSC, andaccelerating leukemogenesis [283]. The differences between LSCs fromdifferent leukemias or different subpopulations of LSCs may reflectcomplex mechanisms by which TGF-β differentially regulates themaintenance of normal HSC subpopulations [244]. Further studies onthe effects of TGF-β on LSC maintenance, and their mechanisms, willyield additional prognostic methods and new therapeutic targets in thetreatment of leukemia.

5.4. Tumor stromal cells

Functional interactions of CSCswith stromal cells that constitute thetumor microenvironment are important for tumor formation, mainte-nance and progression. Most of the tumor microenvironment consistsof cancer-associated fibroblasts (CAFs), which closely resemble myofi-broblasts and express α-smooth muscle actin. Although the origin ofCAFs is not clear, several studies have demonstrated that bonemarrow-derived MSCs and endothelial cells provide a source of CAFs [284,285].TGF-β family signaling has important roles in the interactions betweentumor cells and CAFs. In mouse models of inflammation-dependentgastric cancer, TGF-β induces differentiation of bone marrow-derivedMSCs to myofibroblasts, expansion of these cells and recruitment totumors through induced expression of the secreted chemokine SDF-1αbymyofibroblasts [286]. SDF-1α functions, in an autocrine fashion, as astimulator of proliferation ofmyofibroblasts, and, in a paracrine fashion,as an attractant ofmyofibroblasts to tumor cells,which express the SDF-1α receptor CXCR4. The myofibroblasts are recruited to tumors andsupport cancer cells as CAFs [286]. Additionally, inhibition of TGF-βsignaling or CXCR4 signaling reduces tumor growth in vivo [286]. On theother hand, cancer-associated MSCs isolated from ovarian cancerascites show increased expression of BMP2, 4 and 6, when comparedwith normalMSCs. BMPs expressed by cancer-associatedMSCs increasethe number of human ovarian CSCs and promote tumorigenesis [287].TGF-β also stimulates the generation of CAFs from endothelial cells,which occurs through endothelial to mesenchymal transition [285].Although various findings suggest the importance of TGF-β familysignaling in the interactions of CSCs and CAFs, little is known about theroles of TGF-β family signaling in the regulation of tumor stromal cells.Further studies on the interaction of CSCs and tumor stromal cells mayreveal additional roles of TGF-β family signaling in the maintenance ofCSCs, through effects on tumor stromal cells.

6. Conclusions and future perspective

CSCs, and embryonic and somatic stem cells have striking similar-ities in their responses to signals from the surrounding microenviron-ment, yet some aspects are distinct. Since TGF-β family signalingappears to have different effects on subpopulations of certain tissue


stem cells, such as intestinal stem cells or hematopoietic stem cells, it isto be expected that diverse effects of TGF-β family signaling are seen indifferent types of CSCs. According to their innate properties andacquired changes, several different approaches to target these sub-populations are needed. Further mechanistic studies on TGF-β familysignal transduction and the cross-talk with other signaling pathways indifferent subpopulations of cancer and normal stem cells will provide abetter understanding of the roles of TGF-β family signaling in theregulation of stem cells. These basic studiesmay lead to development ofa new therapeutic or prognostic strategy for the treatment of cancers.

References

[1] G.R. Martin, Isolation of a pluripotent cell line from early mouse embryoscultured in medium conditioned by teratocarcinoma stem cells, Proc. Natl. Acad.Sci. U. S. A. 78 (1981) 7634–7638.

[2] M.J. Evans, M.H. Kaufman, Establishment in culture of pluripotential cells frommouse embryos, Nature 292 (1981) 154–156.

[3] J.A. Thomson, J. Itskovitz-Eldor, S.S. Shapiro, M.A. Waknitz, J.J. Swiergiel, V.S.Marshall, J.M. Jones, Embryonic stem cell lines derived from human blastocysts,Science 282 (1998) 1145–1147.

[4] C.A. Cowan, I. Klimanskaya, J. McMahon, J. Atienza, J. Witmyer, J.P. Zucker, S.Wang, C.C. Morton, A.P. McMahon, D. Powers, D.A. Melton, Derivation ofembryonic stem-cell lines from human blastocysts, N. Engl. J. Med. 350 (2004)1353–1356.

[5] I.G. Brons, L.E. Smithers, M.W. Trotter, P. Rugg-Gunn, B. Sun, S.M. Chuva de SousaLopes, S.K. Howlett, A. Clarkson, L. Ahrlund-Richter, R.A. Pedersen, L. Vallier,Derivation of pluripotent epiblast stem cells from mammalian embryos, Nature448 (2007) 191–195.

[6] P.J. Tesar, J.G. Chenoweth, F.A. Brook, T.J. Davies, E.P. Evans, D.L. Mack, R.L.Gardner, R.D. McKay, New cell lines frommouse epiblast share defining featureswith human embryonic stem cells, Nature 448 (2007) 196–199.

[7] H. Weintraub, S.J. Tapscott, R.L. Davis, M.J. Thayer, M.A. Adam, A.B. Lassar, A.D.Miller, Activation of muscle-specific genes in pigment, nerve, fat, liver, andfibroblast cell lines by forced expression of MyoD, Proc. Natl. Acad. Sci. U. S. A. 86(1989) 5434–5438.

[8] P.J. Miettinen, R. Ebner, A.R. Lopez, R. Derynck, TGF-beta induced transdiffer-entiation of mammary epithelial cells to mesenchymal cells: involvement oftype I receptors, J. Cell Biol. 127 (1994) 2021–2036.

[9] H. Kulessa, J. Frampton, T. Graf, GATA-1 reprograms avian myelomonocytic celllines into eosinophils, thromboblasts, and erythroblasts, Genes Dev. 9 (1995)1250–1262.

[10] H. Xie, M. Ye, R. Feng, T. Graf, Stepwise reprogramming of B cells intomacrophages, Cell 117 (2004) 663–676.

[11] S.C. Bendall, M.H. Stewart, P. Menendez, D. George, K. Vijayaragavan, T.Werbowetski-Ogilvie, V. Ramos-Mejia, A. Rouleau, J. Yang, M. Bosse, G. Lajoie,M. Bhatia, IGF and FGF cooperatively establish the regulatory stem cell niche ofpluripotent human cells in vitro, Nature 448 (2007) 1015–1021.

[12] F. Ferraro, C.L. Celso, D. Scadden, Adult stem cells and their niches, Adv. Exp.Med. Biol. 695 (2010) 155–168.

[13] J. Voog, D.L. Jones, Stem cells and the niche: a dynamic duo, Cell Stem Cell 6(2010) 103–115.

[14] H.J. Snippert, H. Clevers, Tracking adult stem cells, EMBO Rep. 12 (2011)113–122.

[15] K. Takahashi, S. Yamanaka, Induction of pluripotent stem cells from mouseembryonic and adult fibroblast cultures by defined factors, Cell 126 (2006)663–676.

[16] K. Takahashi, K. Tanabe, M. Ohnuki, M. Narita, T. Ichisaka, K. Tomoda, S.Yamanaka, Induction of pluripotent stem cells from adult human fibroblasts bydefined factors, Cell 131 (2007) 861–872.

[17] J. Yu, M.A. Vodyanik, K. Smuga-Otto, J. Antosiewicz-Bourget, J.L. Frane, S. Tian, J.Nie, G.A. Jonsdottir, V. Ruotti, R. Stewart, I.I. Slukvin, J.A. Thomson, Inducedpluripotent stem cell lines derived from human somatic cells, Science 318(2007) 1917–1920.

[18] Q. Zhou, J. Brown, A. Kanarek, J. Rajagopal, D.A. Melton, In vivo reprogrammingof adult pancreatic exocrine cells to beta-cells, Nature 455 (2008) 627–632.

[19] S. Kajimura, P. Seale, K. Kubota, E. Lunsford, J.V. Frangioni, S.P. Gygi, B.M.Spiegelman, Initiation of myoblast to brown fat switch by a PRDM16-C/EBP-betatranscriptional complex, Nature 460 (2009) 1154–1158.

[20] M. Ieda, J.D. Fu, P. Delgado-Olguin, V. Vedantham, Y. Hayashi, B.G. Bruneau, D.Srivastava, Direct reprogramming of fibroblasts into functional cardiomyocytesby defined factors, Cell 142 (2010) 375–386.

[21] P. Li, S. Burke, J. Wang, X. Chen, M. Ortiz, S.C. Lee, D. Lu, L. Campos, D. Goulding,B.L. Ng, G. Dougan, B. Huntly, B. Gottgens, N.A. Jenkins, N.G. Copeland, F. Colucci,P. Liu, Reprogramming of T cells to natural killer-like cells upon Bcl11b deletion,Science 329 (2010) 85–89.

[22] T. Vierbuchen, A. Ostermeier, Z.P. Pang, Y. Kokubu, T.C. Sudhof, M. Wernig,Direct conversion of fibroblasts to functional neurons by defined factors, Nature463 (2010) 1035–1041.

[23] T. Reya, S.J. Morrison, M.F. Clarke, I.L. Weissman, Stem cells, cancer, and cancerstem cells, Nature 414 (2001) 105–111.

[24] J.E. Visvader, G.J. Lindeman, Cancer stem cells in solid tumours: accumulatingevidence and unresolved questions, Nat. Rev. Cancer 8 (2008) 755–768.

[25] J.E. Visvader, Cells of origin in cancer, Nature 469 (2011) 314–322.[26] S.W. Lane, D.T. Scadden, D.G. Gilliland, The leukemic stem cell niche: current

concepts and therapeutic opportunities, Blood 114 (2009) 1150–1157.[27] D. James, A.J. Levine, D. Besser, A. Hemmati-Brivanlou, TGFbeta/activin/nodal

signaling is necessary for the maintenance of pluripotency in human embryonicstem cells, Development 132 (2005) 1273–1282.

[28] N. Oshimori, E. Fuchs, Paracrine TGF-beta signaling counterbalances BMP-mediated repression in hair follicle stem cell activation, Cell Stem Cell 10 (2012)63–75.

[29] E.K. Nishimura, M. Suzuki, V. Igras, J. Du, S. Lonning, Y. Miyachi, J. Roes, F.Beermann, D.E. Fisher, Key roles for transforming growth factor beta inmelanocyte stem cell maintenance, Cell Stem Cell 6 (2010) 130–140.

[30] S.A. Mani, W. Guo, M.J. Liao, E.N. Eaton, A. Ayyanan, A.Y. Zhou, M. Brooks, F.Reinhard, C.C. Zhang, M. Shipitsin, L.L. Campbell, K. Polyak, C. Brisken, J. Yang,R.A. Weinberg, The epithelial–mesenchymal transition generates cells withproperties of stem cells, Cell 133 (2008) 704–715.

[31] C. Scheel, E.N. Eaton, S.H. Li, C.L. Chaffer, F. Reinhardt, K.J. Kah, G. Bell, W. Guo, J.Rubin, A.L. Richardson, R.A. Weinberg, Paracrine and autocrine signals induceand maintain mesenchymal and stem cell states in the breast, Cell 145 (2011)926–940.

[32] J.P. Annes, J.S. Munger, D.B. Rifkin, Making sense of latent TGFbeta activation,J. Cell Sci. 116 (2003) 217–224.

[33] R. Derynck, K. Miyazono, TGF-beta and the TGF-beta family, in: R. Derynck, K.Miyazono (Eds.), The TGF-beta family, Cold Spring Harbor Laboratory Press,New York, 2008, pp. 29–44.

[34] Y. Shi, J. Massague, Mechanisms of TGF-beta signaling from cell membrane tothe nucleus, Cell 113 (2003) 685–700.

[35] M.Y. Wu, C.S. Hill, TGF-beta superfamily signaling in embryonic developmentand homeostasis, Dev. Cell 16 (2009) 329–343.

[36] M.M. Shen, Nodal signaling: developmental roles and regulation, Development134 (2007) 1023–1034.

[37] R. Derynck, R.J. Akhurst, Differentiation plasticity regulated by TGF-beta familyproteins in development and disease, Nat. Cell Biol. 9 (2007) 1000–1004.

[38] M. Whitman, Smads and early developmental signaling by the TGFbetasuperfamily, Genes Dev. 12 (1998) 2445–2462.

[39] C.H. Heldin, K. Miyazono, P. ten Dijke, TGF-beta signalling from cell membraneto nucleus through SMAD proteins, Nature 390 (1997) 465–471.

[40] J. Massague, J. Seoane, D. Wotton, Smad transcription factors, Genes Dev. 19(2005) 2783–2810.

[41] R. Derynck, Y.E. Zhang, Smad-dependent and Smad-independent pathways inTGF-beta family signalling, Nature 425 (2003) 577–584.

[42] X.H. Feng, R. Derynck, Specificity and versatility in TGF-beta signaling throughSmads, Annu. Rev. Cell Dev. Biol. 21 (2005) 659–693.

[43] J. Massague, D. Wotton, Transcriptional control by the TGF-beta/Smad signalingsystem, EMBO J. 19 (2000) 1745–1754.

[44] X. Chen, E. Weisberg, V. Fridmacher, M. Watanabe, G. Naco, M. Whitman, Smad4and FAST-1 in the assembly of activin-responsive factor, Nature 389 (1997)85–89.

[45] E. Labbe, C. Silvestri, P.A. Hoodless, J.L. Wrana, L. Attisano, Smad2 and Smad3positively and negatively regulate TGF beta-dependent transcription throughthe forkhead DNA-binding protein FAST2, Mol. Cell 2 (1998) 109–120.

[46] T. Alliston, L. Choy, P. Ducy, G. Karsenty, R. Derynck, TGF-beta-inducedrepression of CBFA1 by Smad3 decreases cbfa1 and osteocalcin expression andinhibits osteoblast differentiation, EMBO J. 20 (2001) 2254–2272.

[47] D. Liu, B.L. Black, R. Derynck, TGF-beta inhibits muscle differentiation throughfunctional repression of myogenic transcription factors by Smad3, Genes Dev. 15(2001) 2950–2966.

[48] L. Choy, R. Derynck, Transforming growth factor-beta inhibits adipocyte differen-tiation by Smad3 interacting with CCAAT/enhancer-binding protein (C/EBP) andrepressing C/EBP transactivation function, J. Biol. Chem. 278 (2003) 9609–9619.

[49] A. Suzuki, A. Raya, Y. Kawakami, M. Morita, T. Matsui, K. Nakashima, F.H. Gage, C.Rodriguez-Esteban, J.C. Izpisua Belmonte, Nanog binds to Smad1 and blocksbone morphogenetic protein-induced differentiation of embryonic stem cells,Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 10294–10299.

[50] R.H. Xu, T.L. Sampsell-Barron, F. Gu, S. Root, R.M. Peck, G. Pan, J. Yu, J.Antosiewicz-Bourget, S. Tian, R. Stewart, J.A. Thomson, NANOG is a direct targetof TGFbeta/activin-mediated SMAD signaling in human ESCs, Cell Stem Cell 3(2008) 196–206.

[51] L. Vallier, S. Mendjan, S. Brown, Z. Chng, A. Teo, L.E. Smithers, M.W. Trotter, C.H.Cho, A. Martinez, P. Rugg-Gunn, G. Brons, R.A. Pedersen, Activin/Nodal signallingmaintains pluripotency by controlling Nanog expression, Development 136(2009) 1339–1349.

[52] A.C. Mullen, D.A. Orlando, J.J. Newman, J. Loven, R.M. Kumar, S. Bilodeau, J. Reddy,M.G. Guenther, R.P. DeKoter, R.A. Young, Master transcription factors determinecell-type-specific responses to TGF-beta signaling, Cell 147 (2011) 565–576.

[53] S.J. Yoon, A.E. Wills, E. Chuong, R. Gupta, J.C. Baker, HEB and E2A function asSMAD/FOXH1 cofactors, Genes Dev. 25 (2011) 1654–1661.

[54] Q. Xi, Z. Wang, A.I. Zaromytidou, X.H. Zhang, L.F. Chow-Tsang, J.X. Liu, H. Kim, A.Barlas, K. Manova-Todorova, V. Kaartinen, L. Studer, W. Mark, D.J. Patel, J.Massague, A poised chromatin platform for TGF-beta access to masterregulators, Cell 147 (2011) 1511–1524.

[55] O. Dahle, A. Kumar, M.R. Kuehn, Nodal signaling recruits the histonedemethylase Jmjd3 to counteract polycomb-mediated repression at targetgenes, Sci. Signal. 3 (2010) ra48.


[56] A. Moustakas, C.H. Heldin, Non-Smad TGF-beta signals, J. Cell Sci. 118 (2005)3573–3584.

[57] Y.E. Zhang, Non-Smad pathways in TGF-beta signaling, Cell Res. 19 (2009)128–139.

[58] M.K. Lee, C. Pardoux, M.C. Hall, P.S. Lee, D. Warburton, J. Qing, S.M. Smith, R.Derynck, TGF-beta activates Erk MAP kinase signalling through direct phos-phorylation of ShcA, EMBO J. 26 (2007) 3957–3967.

[59] A. Sorrentino, N. Thakur, S. Grimsby, A. Marcusson, V. von Bulow, N. Schuster, S.Zhang, C.H. Heldin, M. Landstrom, The type I TGF-beta receptor engages TRAF6to activate TAK1 in a receptor kinase-independent manner, Nat. Cell Biol. 10(2008) 1199–1207.

[60] M. Yamashita, K. Fatyol, C. Jin, X. Wang, Z. Liu, Y.E. Zhang, TRAF6 mediatesSmad-independent activation of JNK and p38 by TGF-beta, Mol. Cell 31 (2008)918–924.

[61] N.A. Bhowmick, M. Ghiassi, A. Bakin, M. Aakre, C.A. Lundquist, M.E. Engel, C.L.Arteaga, H.L. Moses, Transforming growth factor-beta1 mediates epithelial tomesenchymal transdifferentiation through a RhoA-dependent mechanism, Mol.Biol. Cell 12 (2001) 27–36.

[62] S. Edlund, M. Landstrom, C.H. Heldin, P. Aspenstrom, Transforming growthfactor-beta-induced mobilization of actin cytoskeleton requires signaling bysmall GTPases Cdc42 and RhoA, Mol. Biol. Cell 13 (2002) 902–914.

[63] B. Ozdamar, R. Bose, M. Barrios-Rodiles, H.R. Wang, Y. Zhang, J.L. Wrana,Regulation of the polarity protein Par6 by TGFbeta receptors controls epithelialcell plasticity, Science 307 (2005) 1603–1609.

[64] L. Vardouli, A. Moustakas, C. Stournaras, LIM-kinase 2 and cofilin phosphoryla-tion mediate actin cytoskeleton reorganization induced by transforming growthfactor-beta, J. Biol. Chem. 280 (2005) 11448–11457.

[65] A.V. Bakin, A.K. Tomlinson, N.A. Bhowmick, H.L. Moses, C.L. Arteaga, Phos-phatidylinositol 3-kinase function is required for transforming growth factorbeta-mediated epithelial to mesenchymal transition and cell migration, J. Biol.Chem. 275 (2000) 36803–36810.

[66] I. Shin, A.V. Bakin, U. Rodeck, A. Brunet, C.L. Arteaga, Transforming growth factorbeta enhances epithelial cell survival via Akt-dependent regulation of FKHRL1,Mol. Biol. Cell 12 (2001) 3328–3339.

[67] S. Lamouille, R. Derynck, Cell size and invasion in TGF-beta-induced epithelial tomesenchymal transition is regulated by activation of the mTOR pathway, J. CellBiol. 178 (2007) 437–451.

[68] S. Lamouille, E. Connolly, J.W. Smyth, R.J. Akhurst, R. Derynck, TGF-beta-inducedactivation of mTOR complex 2 drives epithelial–mesenchymal transition andcell invasion, J. Cell Sci. 125 (2012) 1259–1273.

[69] K. Okita, T. Ichisaka, S. Yamanaka, Generation of germline-competent inducedpluripotent stem cells, Nature 448 (2007) 313–317.

[70] K. Kim, A. Doi, B.Wen, K. Ng, R. Zhao, P. Cahan, J. Kim,M.J. Aryee, H. Ji, L.I. Ehrlich, A.Yabuuchi, A. Takeuchi, K.C. Cunniff, H. Hongguang, S. McKinney-Freeman, O.Naveiras, T.J. Yoon, R.A. Irizarry, N. Jung, J. Seita, J. Hanna, P. Murakami, R. Jaenisch,R. Weissleder, S.H. Orkin, I.L. Weissman, A.P. Feinberg, G.Q. Daley, Epigeneticmemory in induced pluripotent stem cells, Nature 467 (2010) 285–290.

[71] J.M. Polo, S. Liu, M.E. Figueroa, W. Kulalert, S. Eminli, K.Y. Tan, E. Apostolou, M.Stadtfeld, Y. Li, T. Shioda, S. Natesan, A.J. Wagers, A. Melnick, T. Evans, K.Hochedlinger, Cell type of origin influences the molecular and functionalproperties of mouse induced pluripotent stem cells, Nat. Biotechnol. 28 (2010)848–855.

[72] R. Lister, M. Pelizzola, Y.S. Kida, R.D. Hawkins, J.R. Nery, G. Hon, J.Antosiewicz-Bourget, R. O'Malley, R. Castanon, S. Klugman, M. Downes, R. Yu,R. Stewart, B. Ren, J.A. Thomson, R.M. Evans, J.R. Ecker, Hotspots of aberrantepigenomic reprogramming in human induced pluripotent stem cells, Nature471 (2011) 68–73.

[73] D.H. Phanstiel, J. Brumbaugh, C.D. Wenger, S. Tian, M.D. Probasco, D.J. Bailey, D.L.Swaney, M.A. Tervo, J.M. Bolin, V. Ruotti, R. Stewart, J.A. Thomson, J.J. Coon,Proteomic and phosphoproteomic comparison of human ES and iPS cells, Nat.Methods 8 (2011) 821–827.

[74] J.H. Hanna, K. Saha, R. Jaenisch, Pluripotency and cellular reprogramming: facts,hypotheses, unresolved issues, Cell 143 (2010) 508–525.

[75] B. Greber, G. Wu, C. Bernemann, J.Y. Joo, D.W. Han, K. Ko, N. Tapia, D. Sabour, J.Sterneckert, P. Tesar, H.R. Scholer, Conserved and divergent roles of FGFsignaling in mouse epiblast stem cells and human embryonic stem cells, CellStem Cell 6 (2010) 215–226.

[76] L. Vallier, T. Touboul, Z. Chng, M. Brimpari, N. Hannan, E. Millan, L.E. Smithers, M.Trotter, P. Rugg-Gunn, A. Weber, R.A. Pedersen, Early cell fate decisions ofhuman embryonic stem cells and mouse epiblast stem cells are controlled by thesame signalling pathways, PLoS One 4 (2009) e6082.

[77] R.H. Xu, X. Chen, D.S. Li, R. Li, G.C. Addicks, C. Glennon, T.P. Zwaka, J.A. Thomson,BMP4 initiates human embryonic stem cell differentiation to trophoblast, Nat.Biotechnol. 20 (2002) 1261–1264.

[78] A.S. Bernardo, T. Faial, L. Gardner, K.K. Niakan, D. Ortmann, C.E. Senner, E.M.Callery, M.W. Trotter, M. Hemberger, J.C. Smith, L. Bardwell, A. Moffett, R.A.Pedersen, BRACHYURY and CDX2 mediate BMP-induced differentiation ofhuman and mouse pluripotent stem cells into embryonic and extraembryoniclineages, Cell Stem Cell 9 (2011) 144–155.

[79] V. Tropepe, S. Hitoshi, C. Sirard, T.W. Mak, J. Rossant, D. van der Kooy, Directneural fate specification from embryonic stem cells: a primitive mammalianneural stem cell stage acquired through a default mechanism, Neuron 30 (2001)65–78.

[80] Q.L. Ying, J. Nichols, I. Chambers, A. Smith, BMP induction of Id proteinssuppresses differentiation and sustains embryonic stem cell self-renewal incollaboration with STAT3, Cell 115 (2003) 281–292.

[81] X. Qi, T.G. Li, J. Hao, J. Hu, J. Wang, H. Simmons, S. Miura, Y. Mishina, G.Q. Zhao,BMP4 supports self-renewal of embryonic stem cells by inhibiting mitogen-activated protein kinase pathways, Proc. Natl. Acad. Sci. U. S. A. 101 (2004)6027–6032.

[82] G. Guo, J. Yang, J. Nichols, J.S. Hall, I. Eyres, W. Mansfield, A. Smith, Klf4 revertsdevelopmentally programmed restriction of ground state pluripotency, Devel-opment 136 (2009) 1063–1069.

[83] K. Ogawa, A. Saito, H. Matsui, H. Suzuki, S. Ohtsuka, D. Shimosato, Y. Morishita, T.Watabe, H. Niwa, K. Miyazono, Activin–Nodal signaling is involved inpropagation of mouse embryonic stem cells, J. Cell Sci. 120 (2007) 55–65.

[84] S. Bao, F. Tang, X. Li, K. Hayashi, A. Gillich, K. Lao, M.A. Surani, Epigeneticreversion of post-implantation epiblast to pluripotent embryonic stem cells,Nature 461 (2009) 1292–1295.

[85] J. Nichols, A. Smith, Naive and primed pluripotent states, Cell Stem Cell 4 (2009)487–492.

[86] H. Zhou, W. Li, S. Zhu, J.Y. Joo, J.T. Do, W. Xiong, J.B. Kim, K. Zhang, H.R. Scholer, S.Ding, Conversion of mouse epiblast stem cells to an earlier pluripotency state bysmall molecules, J. Biol. Chem. 285 (2010) 29676–29680.

[87] J. Yang, A.L. van Oosten, T.W. Theunissen, G. Guo, J.C. Silva, A. Smith, Stat3activation is limiting for reprogramming to ground state pluripotency, Cell StemCell 7 (2010) 319–328.

[88] Y. Xu, X. Zhu, H.S. Hahm, W. Wei, E. Hao, A. Hayek, S. Ding, Revealing a coresignaling regulatory mechanism for pluripotent stem cell survival andself-renewal by small molecules, Proc. Natl. Acad. Sci. U. S. A. 107 (2010)8129–8134.

[89] W. Li, W. Wei, S. Zhu, J. Zhu, Y. Shi, T. Lin, E. Hao, A. Hayek, H. Deng, S. Ding,Generation of rat and human induced pluripotent stem cells by combininggenetic reprogramming and chemical inhibitors, Cell Stem Cell 4 (2009) 16–19.

[90] J. Hanna, A.W. Cheng, K. Saha, J. Kim, C.J. Lengner, F. Soldner, J.P. Cassady, J.Muffat, B.W. Carey, R. Jaenisch, Human embryonic stem cells with biological andepigenetic characteristics similar to those of mouse ESCs, Proc. Natl. Acad. Sci. U. S. A.107 (2010) 9222–9227.

[91] A.M. Singh, T. Hamazaki, K.E. Hankowski, N. Terada, A heterogeneous expressionpattern for Nanog in embryonic stem cells, Stem Cells 25 (2007) 2534–2542.

[92] K. Hayashi, S.M. Lopes, F. Tang, M.A. Surani, Dynamic equilibrium andheterogeneity of mouse pluripotent stem cells with distinct functional andepigenetic states, Cell Stem Cell 3 (2008) 391–401.

[93] D.W. Han, N. Tapia, J.Y. Joo, B. Greber, M.J. Arauzo-Bravo, C. Bernemann, K. Ko, G.Wu, M. Stehling, J.T. Do, H.R. Scholer, Epiblast stem cell subpopulationsrepresent mouse embryos of distinct pregastrulation stages, Cell 143 (2010)617–627.

[94] M.H. Stewart, M. Bosse, K. Chadwick, P. Menendez, S.C. Bendall, M. Bhatia, Clonalisolation of hESCs reveals heterogeneity within the pluripotent stem cellcompartment, Nat. Methods 3 (2006) 807–815.

[95] X. Xu, V.L. Browning, J.S. Odorico, Activin, BMP and FGF pathways cooperate topromote endoderm and pancreatic lineage cell differentiation from humanembryonic stem cells, Mech. Dev. 128 (2011) 412–427.

[96] S.J. Kattman, A.D. Witty, M. Gagliardi, N.C. Dubois, M. Niapour, A. Hotta, J. Ellis, G.Keller, Stage-specific optimization of activin/nodal and BMP signaling promotescardiac differentiation of mouse and human pluripotent stem cell lines, CellStem Cell 8 (2011) 228–240.

[97] T. Sumi, N. Tsuneyoshi, N. Nakatsuji, H. Suemori, Defining early lineagespecification of human embryonic stem cells by the orchestrated balance ofcanonical Wnt/beta-catenin, Activin/Nodal and BMP signaling, Development135 (2008) 2969–2979.

[98] P. Gadue, T.L. Huber, P.J. Paddison, G.M. Keller, Wnt and TGF-beta signaling arerequired for the induction of an in vitro model of primitive streak formationusing embryonic stem cells, Proc. Natl. Acad. Sci. U. S. A. 103 (2006)16806–16811.

[99] K.A. D'Amour, A.D. Agulnick, S. Eliazer, O.G. Kelly, E. Kroon, E.E. Baetge, Efficientdifferentiation of human embryonic stem cells to definitive endoderm, Nat.Biotechnol. 23 (2005) 1534–1541.

[100] M. Yasunaga, S. Tada, S. Torikai-Nishikawa, Y. Nakano, M. Okada, L.M. Jakt, S.Nishikawa, T. Chiba, T. Era, Induction and monitoring of definitive and visceralendoderm differentiation of mouse ES cells, Nat. Biotechnol. 23 (2005)1542–1550.

[101] Z. Gu, M. Nomura, B.B. Simpson, H. Lei, A. Feijen, J. van den Eijnden-van Raaij,P.K. Donahoe, E. Li, The type I activin receptor ActRIB is required for egg cylinderorganization and gastrulation in the mouse, Genes Dev. 12 (1998) 844–857.

[102] M.V. Wiles, B.M. Johansson, Embryonic stem cell development in a chemicallydefined medium, Exp. Cell Res. 247 (1999) 241–248.

[103] R.H. Xu, R.M. Peck, D.S. Li, X. Feng, T. Ludwig, J.A. Thomson, Basic FGF andsuppression of BMP signaling sustain undifferentiated proliferation of human EScells, Nat. Methods 2 (2005) 185–190.

[104] A. Kubo, K. Shinozaki, J.M. Shannon, V. Kouskoff, M. Kennedy, S. Woo, H.J.Fehling, G. Keller, Development of definitive endoderm from embryonic stemcells in culture, Development 131 (2004) 1651–1662.

[105] S. Tada, T. Era, C. Furusawa, H. Sakurai, S. Nishikawa, M. Kinoshita, K. Nakao, T.Chiba, Characterization of mesendoderm: a diverging point of the definitiveendoderm and mesoderm in embryonic stem cell differentiation culture,Development 132 (2005) 4363–4374.

[106] T. Fei, S. Zhu, K. Xia, J. Zhang, Z. Li, J.D. Han, Y.G. Chen, Smad2 mediatesActivin/Nodal signaling in mesendoderm differentiation of mouse embryonicstem cells, Cell Res. 20 (2010) 1306–1318.

[107] K.L. Lee, S.K. Lim, Y.L. Orlov, Y. Yit le, H. Yang, L.T. Ang, L. Poellinger, B. Lim, GradedNodal/Activin signaling titrates conversion of quantitative phospho-Smad2


levels into qualitative embryonic stem cell fate decisions, PLoS Genet. 7 (2011)e1002130.

[108] S. Brown, A. Teo, S. Pauklin, N. Hannan, C.H. Cho, B. Lim, L. Vardy, N.R. Dunn, M.Trotter, R. Pedersen, L. Vallier, Activin/Nodal signaling controls divergenttranscriptional networks in human embryonic stem cells and in endodermprogenitors, Stem Cells 29 (2011) 1176–1185.

[109] S.J. Kattman, T.L. Huber, G.M. Keller, Multipotent flk-1+ cardiovascularprogenitor cells give rise to the cardiomyocyte, endothelial, and vascularsmooth muscle lineages, Dev. Cell 11 (2006) 723–732.

[110] S.M. Chambers, C.A. Fasano, E.P. Papapetrou, M. Tomishima, M. Sadelain, L.Studer, Highly efficient neural conversion of human ES and iPS cells by dualinhibition of SMAD signaling, Nat. Biotechnol. 27 (2009) 275–280.

[111] J. Zhou, P. Su, D. Li, S. Tsang, E. Duan, F. Wang, High-efficiency induction ofneural conversion in human ESCs and human induced pluripotent stem cellswith a single chemical inhibitor of transforming growth factor beta superfamilyreceptors, Stem Cells 28 (2010) 1741–1750.

[112] D.W. Han, B. Greber, G. Wu, N. Tapia, M.J. Arauzo-Bravo, K. Ko, C. Bernemann, M.Stehling, H.R. Scholer, Direct reprogramming of fibroblasts into epiblast stemcells, Nat. Cell Biol. 13 (2011) 66–71.

[113] J.K. Ichida, J. Blanchard, K. Lam, E.Y. Son, J.E. Chung, D. Egli, K.M. Loh, A.C. Carter,F.P. Di Giorgio, K. Koszka, D. Huangfu, H. Akutsu, D.R. Liu, L.L. Rubin, K. Eggan, Asmall-molecule inhibitor of TGF-beta signaling replaces sox2 in reprogrammingby inducing nanog, Cell Stem Cell 5 (2009) 491–503.

[114] N. Maherali, K. Hochedlinger, Tgfbeta signal inhibition cooperates in theinduction of iPSCs and replaces Sox2 and cMyc, Curr. Biol. 19 (2009) 1718–1723.

[115] Z. Li, C.S. Yang, K. Nakashima, T.M. Rana, Small RNA-mediated regulation of iPScell generation, EMBO J. 30 (2011) 823–834.

[116] N. Miyoshi, H. Ishii, H. Nagano, N. Haraguchi, D.L. Dewi, Y. Kano, S. Nishikawa, M.Tanemura, K. Mimori, F. Tanaka, T. Saito, J. Nishimura, I. Takemasa, T.Mizushima, M. Ikeda, H. Yamamoto, M. Sekimoto, Y. Doki, M. Mori, Reprogram-ming of mouse and human cells to pluripotency using mature microRNAs, CellStem Cell 8 (2011) 633–638.

[117] D. Subramanyam, S. Lamouille, R.L. Judson, J.Y. Liu, N. Bucay, R. Derynck, R.Blelloch, Multiple targets of miR-302 and miR-372 promote reprogramming ofhuman fibroblasts to induced pluripotent stem cells, Nat. Biotechnol. 29 (2011)443–448.

[118] P. Samavarchi-Tehrani, A. Golipour, L. David, H.K. Sung, T.A. Beyer, A. Datti, K.Woltjen, A. Nagy, J.L. Wrana, Functional genomics reveals a BMP-drivenmesenchymal-to-epithelial transition in the initiation of somatic cell repro-gramming, Cell Stem Cell 7 (2010) 64–77.

[119] R. Li, J. Liang, S. Ni, T. Zhou, X. Qing, H. Li, W. He, J. Chen, F. Li, Q. Zhuang, B. Qin, J.Xu, W. Li, J. Yang, Y. Gan, D. Qin, S. Feng, H. Song, D. Yang, B. Zhang, L. Zeng, L. Lai,M.A. Esteban, D. Pei, A mesenchymal-to-epithelial transition initiates and isrequired for the nuclear reprogramming of mouse fibroblasts, Cell Stem Cell 7(2010) 51–63.

[120] D.A. Card, P.B. Hebbar, L. Li, K.W. Trotter, Y. Komatsu, Y. Mishina, T.K. Archer,Oct4/Sox2-regulated miR-302 targets cyclin D1 in human embryonic stem cells,Mol. Cell. Biol. 28 (2008) 6426–6438.

[121] A. Rosa, A.H. Brivanlou, A regulatory circuitry comprised of miR-302 and thetranscription factors OCT4 and NR2F2 regulates human embryonic stem celldifferentiation, EMBO J. 30 (2011) 237–248.

[122] I. Lipchina, Y. Elkabetz, M. Hafner, R. Sheridan, A. Mihailovic, T. Tuschl, C. Sander,L. Studer, D. Betel, Genome-wide identification of microRNA targets in human EScells reveals a role for miR-302 in modulating BMP response, Genes Dev. 25(2011) 2173–2186.

[123] A. Barroso-delJesus, G. Lucena-Aguilar, L. Sanchez, G. Ligero, I. Gutierrez-Aranda,P. Menendez, The Nodal inhibitor Lefty is negatively modulated by themicroRNA miR-302 in human embryonic stem cells, FASEB J. 25 (2011)1497–1508.

[124] C.E. Rogler, L. Levoci, T. Ader, A. Massimi, T. Tchaikovskaya, R. Norel, L.E. Rogler,MicroRNA-23b cluster microRNAs regulate transforming growth factor-beta/bonemorphogenetic protein signaling and liver stem cell differentiation by targetingSmads, Hepatology 50 (2009) 575–584.

[125] Q. Wang, Z. Huang, H. Xue, C. Jin, X.L. Ju, J.D. Han, Y.G. Chen, MicroRNA miR-24inhibits erythropoiesis by targeting activin type I receptor ALK4, Blood 111(2008) 588–595.

[126] Y.J. Kim, S.J. Hwang, Y.C. Bae, J.S. Jung, MiR-21 regulates adipogenic differenti-ation through the modulation of TGF-beta signaling in mesenchymal stem cellsderived from human adipose tissue, Stem Cells 27 (2009) 3093–3102.

[127] G. Dontu, W.M. Abdallah, J.M. Foley, K.W. Jackson, M.F. Clarke, M.J. Kawamura,M.S. Wicha, In vitro propagation and transcriptional profiling of humanmammary stem/progenitor cells, Genes Dev. 17 (2003) 1253–1270.

[128] J. Stingl, P. Eirew, I. Ricketson, M. Shackleton, F. Vaillant, D. Choi, H.I. Li, C.J. Eaves,Purification and unique properties of mammary epithelial stem cells, Nature 439(2006) 993–997.

[129] M. Shackleton, F. Vaillant, K.J. Simpson, J. Stingl, G.K. Smyth, M.L. Asselin-Labat,L. Wu, G.J. Lindeman, J.E. Visvader, Generation of a functional mammary glandfrom a single stem cell, Nature 439 (2006) 84–88.

[130] S. Liu, G. Dontu, M.S. Wicha, Mammary stem cells, self-renewal pathways, andcarcinogenesis, Breast Cancer Res. 7 (2005) 86–95.

[131] E.C. Kordon, R.A. McKnight, C. Jhappan, L. Hennighausen, G. Merlino, G.H. Smith,Ectopic TGF beta 1 expression in the secretory mammary epithelium inducesearly senescence of the epithelial stem cell population, Dev. Biol. 168 (1995)47–61.

[132] H. Moses, M.H. Barcellos-Hoff, TGF-beta biology in mammary development andbreast cancer, Cold Spring Harb. Perspect. Biol. 3 (2011) a003277.

[133] J.P. Thiery, H. Acloque, R.Y. Huang, M.A. Nieto, Epithelial–mesenchymaltransitions in development and disease, Cell 139 (2009) 871–890.

[134] R. Kalluri, R.A. Weinberg, The basics of epithelial–mesenchymal transition,J. Clin. Invest. 119 (2009) 1420–1428.

[135] M. Oft, J. Peli, C. Rudaz, H. Schwarz, H. Beug, E. Reichmann, TGF-beta1 andHa-Ras collaborate in modulating the phenotypic plasticity and invasiveness ofepithelial tumor cells, Genes Dev. 10 (1996) 2462–2477.

[136] E. Piek, A. Moustakas, A. Kurisaki, C.H. Heldin, P. ten Dijke, TGF-(beta) type Ireceptor/ALK-5 and Smad proteins mediate epithelial to mesenchymal transdiffer-entiation in NMuMG breast epithelial cells, J. Cell Sci. 112 (Pt 24) (1999) 4557–4568.

[137] A. Moustakas, C.H. Heldin, Signaling networks guiding epithelial–mesenchymaltransitions during embryogenesis and cancer progression, Cancer Sci. 98 (2007)1512–1520.

[138] J. Yang, R.A. Weinberg, Epithelial–mesenchymal transition: at the crossroads ofdevelopment and tumor metastasis, Dev. Cell 14 (2008) 818–829.

[139] J. Xu, S. Lamouille, R. Derynck, TGF-beta-induced epithelial to mesenchymaltransition, Cell Res. 19 (2009) 156–172.

[140] A. Singh, J. Settleman, EMT, cancer stem cells and drug resistance: an emergingaxis of evil in the war on cancer, Oncogene 29 (2010) 4741–4751.

[141] L.G. van der Flier, H. Clevers, Stem cells, self-renewal, and differentiation in theintestinal epithelium, Annu. Rev. Physiol. 71 (2009) 241–260.

[142] N. Barker, J.H. van Es, J. Kuipers, P. Kujala, M. van den Born, M. Cozijnsen, A.Haegebarth, J. Korving, H. Begthel, P.J. Peters, H. Clevers, Identification of stemcells in small intestine and colon by marker gene Lgr5, Nature 449 (2007)1003–1007.

[143] E. Sangiorgi, M.R. Capecchi, Bmi1 is expressed in vivo in intestinal stem cells, Nat.Genet. 40 (2008) 915–920.

[144] N. Takeda, R. Jain, M.R. LeBoeuf, Q. Wang, M.M. Lu, J.A. Epstein, Interconversionbetween intestinal stem cell populations in distinct niches, Science 334 (2011)1420–1424.

[145] H. Tian, B. Biehs, S. Warming, K.G. Leong, L. Rangell, O.D. Klein, F.J. de Sauvage, Areserve stem cell population in small intestine renders Lgr5-positive cellsdispensable, Nature 478 (2011) 255–259.

[146] A. Gregorieff, H. Clevers, Wnt signaling in the intestinal epithelium: fromendoderm to cancer, Genes Dev. 19 (2005) 877–890.

[147] A.P. Haramis, H. Begthel, M. van den Born, J. van Es, S. Jonkheer, G.J. Offerhaus, H.Clevers, De novo crypt formation and juvenile polyposis on BMP inhibition inmouse intestine, Science 303 (2004) 1684–1686.

[148] X.C. He, J. Zhang, W.G. Tong, O. Tawfik, J. Ross, D.H. Scoville, Q. Tian, X. Zeng, X.He, L.M. Wiedemann, Y. Mishina, L. Li, BMP signaling inhibits intestinal stem cellself-renewal through suppression of Wnt–beta-catenin signaling, Nat. Genet. 36(2004) 1117–1121.

[149] J.R. Howe, S. Roth, J.C. Ringold, R.W. Summers, H.J. Jarvinen, P. Sistonen, I.P.Tomlinson, R.S. Houlston, S. Bevan, F.A. Mitros, E.M. Stone, L.A. Aaltonen,Mutations in the SMAD4/DPC4 gene in juvenile polyposis, Science 280 (1998)1086–1088.

[150] J.R. Howe, J.L. Bair, M.G. Sayed, M.E. Anderson, F.A. Mitros, G.M. Petersen, V.E.Velculescu, G. Traverso, B. Vogelstein, Germline mutations of the gene encodingbone morphogenetic protein receptor 1A in juvenile polyposis, Nat. Genet. 28(2001) 184–187.

[151] S.E. Millar, Molecular mechanisms regulating hair follicle development, J. Invest.Dermatol. 118 (2002) 216–225.

[152] R. Schmidt-Ullrich, R. Paus, Molecular principles of hair follicle induction andmorphogenesis, Bioessays 27 (2005) 247–261.

[153] C. Blanpain, E. Fuchs, Epidermal homeostasis: a balancing act of stem cells in theskin, Nat. Rev. Mol. Cell Biol. 10 (2009) 207–217.

[154] G. Cotsarelis, T.T. Sun, R.M. Lavker, Label-retaining cells reside in the bulge areaof pilosebaceous unit: implications for follicular stem cells, hair cycle, and skincarcinogenesis, Cell 61 (1990) 1329–1337.

[155] K. Foitzik, G. Lindner, S. Mueller-Roever, M. Maurer, N. Botchkareva, V.Botchkarev, B. Handjiski, M. Metz, T. Hibino, T. Soma, G.P. Dotto, R. Paus,Control of murine hair follicle regression (catagen) by TGF-beta1 in vivo, FASEBJ. 14 (2000) 752–760.

[156] T. Soma, C.E. Dohrmann, T. Hibino, L.A. Raftery, Profile of transforming growthfactor-beta responses during the murine hair cycle, J. Invest. Dermatol. 121(2003) 969–975.

[157] K. Foitzik, R. Paus, T. Doetschman, G.P. Dotto, The TGF-beta2 isoform is both arequired and sufficient inducer of murine hair follicle morphogenesis, Dev. Biol.212 (1999) 278–289.

[158] K. Kobielak, N. Stokes, J. de la Cruz, L. Polak, E. Fuchs, Loss of a quiescent nichebut not follicle stem cells in the absence of bone morphogenetic proteinsignaling, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 10063–10068.

[159] H. Kulessa, G. Turk, B.L. Hogan, Inhibition of Bmp signaling affects growth anddifferentiation in the anagen hair follicle, EMBO J. 19 (2000) 6664–6674.

[160] V.A. Botchkarev, N.V. Botchkareva, M. Nakamura, O. Huber, K. Funa, R. Lauster, R.Paus, B.A. Gilchrest, Noggin is required for induction of the hair follicle growthphase in postnatal skin, FASEB J. 15 (2001) 2205–2214.

[161] J. Zhang, X.C. He, W.G. Tong, T. Johnson, L.M. Wiedemann, Y. Mishina, J.Q. Feng, L.Li, Bone morphogenetic protein signaling inhibits hair follicle anagen inductionby restricting epithelial stem/progenitor cell activation and expansion, StemCells 24 (2006) 2826–2839.

[162] E.K. Nishimura, S.A. Jordan, H. Oshima, H. Yoshida, M. Osawa, M. Moriyama, I.J.Jackson, Y. Barrandon, Y. Miyachi, S. Nishikawa, Dominant role of the niche inmelanocyte stem-cell fate determination, Nature 416 (2002) 854–860.

[163] E. Steingrimsson, N.G. Copeland, N.A. Jenkins, Melanocytes and the micro-phthalmia transcription factor network, Annu. Rev. Genet. 38 (2004) 365–411.


[164] C. Levy, M. Khaled, D.E. Fisher, MITF: master regulator of melanocytedevelopment and melanoma oncogene, Trends Mol. Med. 12 (2006) 406–414.

[165] E.K. Nishimura, S.R. Granter, D.E. Fisher, Mechanisms of hair graying: incompletemelanocyte stem cell maintenance in the niche, Science 307 (2005) 720–724.

[166] T. Tumbar, G. Guasch, V. Greco, C. Blanpain, W.E. Lowry, M. Rendl, E. Fuchs,Defining the epithelial stem cell niche in skin, Science 303 (2004) 359–363.

[167] S. Tanimura, Y. Tadokoro, K. Inomata, N.T. Binh, W. Nishie, S. Yamazaki, H.Nakauchi, Y. Tanaka, J.R. McMillan, D. Sawamura, K. Yancey, H. Shimizu, E.K.Nishimura, Hair follicle stem cells provide a functional niche for melanocytestem cells, Cell Stem Cell 8 (2011) 177–187.

[168] D.G. Phinney, D.J. Prockop, Concise review: mesenchymal stem/multipotentstromal cells: the state of transdifferentiation and modes of tissue repair—current views, Stem Cells 25 (2007) 2896–2902.

[169] S. Shi, P.M. Bartold, M. Miura, B.M. Seo, P.G. Robey, S. Gronthos, The efficacy ofmesenchymal stem cells to regenerate and repair dental structures, Orthod.Craniofac. Res. 8 (2005) 191–199.

[170] M.F. Pittenger, A.M. Mackay, S.C. Beck, R.K. Jaiswal, R. Douglas, J.D. Mosca, M.A.Moorman, D.W. Simonetti, S. Craig, D.R. Marshak, Multilineage potential of adulthuman mesenchymal stem cells, Science 284 (1999) 143–147.

[171] J.T. Williams, S.S. Southerland, J. Souza, A.F. Calcutt, R.G. Cartledge, Cells isolatedfrom adult human skeletal muscle capable of differentiating into multiplemesodermal phenotypes, Am. Surg. 65 (1999) 22–26.

[172] A.I. Caplan, Mesenchymal stem cells and gene therapy, Clin. Orthop. Relat. Res.(2000) S67–70.

[173] G.M. Boland, G. Perkins, D.J. Hall, R.S. Tuan, Wnt 3a promotes proliferation andsuppresses osteogenic differentiation of adult humanmesenchymal stem cells,J. Cell. Biochem. 93 (2004) 1210–1230.

[174] H. Jian, X. Shen, I. Liu, M. Semenov, X. He, X.F. Wang, Smad3-dependent nucleartranslocation of beta-catenin is required for TGF-beta1-induced proliferation ofbone marrow-derived adult human mesenchymal stem cells, Genes Dev. 20(2006) 666–674.

[175] D. Baksh, G.M. Boland, R.S. Tuan, Cross-talk between Wnt signaling pathways inhuman mesenchymal stem cells leads to functional antagonism duringosteogenic differentiation, J. Cell. Biochem. 101 (2007) 1109–1124.

[176] Z. Liu, Y. Tang, T. Qiu, X. Cao, T.L. Clemens, A dishevelled-1/Smad1 interactioncouples WNT and bone morphogenetic protein signaling pathways in uncom-mitted bone marrow stromal cells, J. Biol. Chem. 281 (2006) 17156–17163.

[177] B.A. Roelen, P. Dijke, Controlling mesenchymal stem cell differentiation byTGFbeta family members, J. Orthop. Sci. 8 (2003) 740–748.

[178] S. Maeda, M. Hayashi, S. Komiya, T. Imamura, K. Miyazono, EndogenousTGF-beta signaling suppresses maturation of osteoblastic mesenchymal cells,EMBO J. 23 (2004) 552–563.

[179] X. Lin, Y.Y. Liang, B. Sun, M. Liang, Y. Shi, F.C. Brunicardi, X.H. Feng, Smad6recruits transcription corepressor CtBP to repress bone morphogeneticprotein-induced transcription, Mol. Cell. Biol. 23 (2003) 9081–9093.

[180] Y.H. Tseng, E. Kokkotou, T.J. Schulz, T.L. Huang, J.N. Winnay, C.M. Taniguchi, T.T.Tran, R. Suzuki, D.O. Espinoza, Y. Yamamoto, M.J. Ahrens, A.T. Dudley, A.W.Norris, R.N. Kulkarni, C.R. Kahn, New role of bone morphogenetic protein 7 inbrown adipogenesis and energy expenditure, Nature 454 (2008) 1000–1004.

[181] G. Ferrari, G. Cusella-De Angelis, M. Coletta, E. Paolucci, A. Stornaiuolo, G. Cossu,F. Mavilio, Muscle regeneration by bone marrow-derived myogenic progenitors,Science 279 (1998) 1528–1530.

[182] K. Goldring, T. Partridge, D. Watt, Muscle stem cells, J. Pathol. 197 (2002)457–467.

[183] D. Orlic, J. Kajstura, S. Chimenti, I. Jakoniuk, S.M. Anderson, B. Li, J. Pickel, R.McKay, B. Nadal-Ginard, D.M. Bodine, A. Leri, P. Anversa, Bone marrow cellsregenerate infarcted myocardium, Nature 410 (2001) 701–705.

[184] C. Toma, M.F. Pittenger, K.S. Cahill, B.J. Byrne, P.D. Kessler, Human mesenchymalstem cells differentiate to a cardiomyocyte phenotype in the adult murine heart,Circulation 105 (2002) 93–98.

[185] J.S. Wang, D. Shum-Tim, J. Galipeau, E. Chedrawy, N. Eliopoulos, R.C. Chiu,Marrow stromal cells for cellular cardiomyoplasty: feasibility and potentialclinical advantages, J. Thorac. Cardiovasc. Surg. 120 (2000) 999–1005.

[186] M.F. Pittenger, B.J. Martin, Mesenchymal stem cells and their potential as cardiactherapeutics, Circ. Res. 95 (2004) 9–20.

[187] J.Y. Min, M.F. Sullivan, Y. Yang, J.P. Zhang, K.L. Converso, J.P. Morgan, Y.F. Xiao,Significant improvement of heart function by cotransplantation of humanmesenchymal stem cells and fetal cardiomyocytes in postinfarcted pigs, Ann.Thorac. Surg. 74 (2002) 1568–1575.

[188] A. Mathur, J.F. Martin, Stem cells and repair of the heart, Lancet 364 (2004)183–192.

[189] C. Stamm, B. Westphal, H.D. Kleine, M. Petzsch, C. Kittner, H. Klinge, C.Schumichen, C.A. Nienaber, M. Freund, G. Steinhoff, Autologous bone-marrowstem-cell transplantation for myocardial regeneration, Lancet 361 (2003)45–46.

[190] Y. Miyahara, N. Nagaya, M. Kataoka, B. Yanagawa, K. Tanaka, H. Hao, K. Ishino, H.Ishida, T. Shimizu, K. Kangawa, S. Sano, T. Okano, S. Kitamura, H. Mori,Monolayered mesenchymal stem cells repair scarred myocardium aftermyocardial infarction, Nat. Med. 12 (2006) 459–465.

[191] K. Kurpinski, H. Lam, J. Chu, A. Wang, A. Kim, E. Tsay, S. Agrawal, D.V. Schaffer, S.Li, Transforming growth factor-beta and notch signaling mediate stem celldifferentiation into smooth muscle cells, Stem Cells 28 (2010) 734–742.

[192] D.R. Campion, The muscle satellite cell: a review, Int. Rev. Cytol. 87 (1984)225–251.

[193] S.B. Charge, M.A. Rudnicki, Cellular and molecular regulation of muscleregeneration, Physiol. Rev. 84 (2004) 209–238.

[194] B. Peault, M. Rudnicki, Y. Torrente, G. Cossu, J.P. Tremblay, T. Partridge, E.Gussoni, L.M. Kunkel, J. Huard, Stem and progenitor cells in skeletal muscledevelopment, maintenance, and therapy, Mol. Ther. 15 (2007) 867–877.

[195] A. Asakura, P. Seale, A. Girgis-Gabardo, M.A. Rudnicki, Myogenic specification ofside population cells in skeletal muscle, J. Cell Biol. 159 (2002) 123–134.

[196] Y. Torrente, M. Belicchi, M. Sampaolesi, F. Pisati, M. Meregalli, G. D'Antona, R.Tonlorenzi, L. Porretti, M. Gavina, K. Mamchaoui, M.A. Pellegrino, D. Furling, V.Mouly, G.S. Butler-Browne, R. Bottinelli, G. Cossu, N. Bresolin, Human circulatingAC133(+) stem cells restore dystrophin expression and ameliorate function indystrophic skeletal muscle, J. Clin. Invest. 114 (2004) 182–195.

[197] B.G. Galvez, M. Sampaolesi, S. Brunelli, D. Covarello, M. Gavina, B. Rossi, G.Constantin, Y. Torrente, G. Cossu, Complete repair of dystrophic skeletal muscleby mesoangioblasts with enhanced migration ability, J. Cell Biol. 174 (2006)231–243.

[198] A. Dellavalle, M. Sampaolesi, R. Tonlorenzi, E. Tagliafico, B. Sacchetti, L. Perani, A.Innocenzi, B.G. Galvez, G. Messina, R. Morosetti, S. Li, M. Belicchi, G. Peretti, J.S.Chamberlain, W.E. Wright, Y. Torrente, S. Ferrari, P. Bianco, G. Cossu, Pericytes ofhuman skeletal muscle are myogenic precursors distinct from satellite cells, Nat.Cell Biol. 9 (2007) 255–267.

[199] E. Negroni, I. Riederer, S. Chaouch, M. Belicchi, P. Razini, J. Di Santo, Y. Torrente,G.S. Butler-Browne, V. Mouly, In vivo myogenic potential of human CD133+muscle-derived stem cells: a quantitative study, Mol. Ther. 17 (2009)1771–1778.

[200] S. Kuang, K. Kuroda, F. Le Grand, M.A. Rudnicki, Asymmetric self-renewal andcommitment of satellite stem cells in muscle, Cell 129 (2007) 999–1010.

[201] T.J. Hawke, D.J. Garry, Myogenic satellite cells: physiology to molecular biology,J. Appl. Physiol. 91 (2001) 534–551.

[202] H.D. Kollias, J.C. McDermott, Transforming growth factor-beta and myostatinsignaling in skeletal muscle, J. Appl. Physiol. 104 (2008) 579–587.

[203] R.E. Allen, L.K. Boxhorn, Regulation of skeletal muscle satellite cell proliferationand differentiation by transforming growth factor-beta, insulin-like growthfactor I, and fibroblast growth factor, J. Cell. Physiol. 138 (1989) 311–315.

[204] D.R. Cook, M.E. Doumit, R.A. Merkel, Transforming growth factor-beta, basicfibroblast growth factor, and platelet-derived growth factor-BB interact to affectproliferation of clonally derived porcine satellite cells, J. Cell. Physiol. 157 (1993)307–312.

[205] J.R. Florini, A.B. Roberts, D.Z. Ewton, S.L. Falen, K.C. Flanders, M.B. Sporn,Transforming growth factor-beta. A very potent inhibitor of myoblast differen-tiation, identical to the differentiation inhibitor secreted by Buffalo rat liver cells,J. Biol. Chem. 261 (1986) 16509–16513.

[206] E.A. Greene, R.E. Allen, Growth factor regulation of bovine satellite cell growth invitro, J. Anim. Sci. 69 (1991) 146–152.

[207] J. Massague, S. Cheifetz, T. Endo, B. Nadal-Ginard, Type beta transforming growthfactor is an inhibitor of myogenic differentiation, Proc. Natl. Acad. Sci. U. S. A. 83(1986) 8206–8210.

[208] E.N. Olson, E. Sternberg, J.S. Hu, G. Spizz, C. Wilcox, Regulation of myogenicdifferentiation by type beta transforming growth factor, J. Cell Biol. 103 (1986)1799–1805.

[209] H. Amthor, R. Macharia, R. Navarrete, M. Schuelke, S.C. Brown, A. Otto, T. Voit, F.Muntoni, G. Vrbova, T. Partridge, P. Zammit, L. Bunger, K. Patel, Lack of myostatinresults in excessive muscle growth but impaired force generation, Proc. Natl.Acad. Sci. U. S. A. 104 (2007) 1835–1840.

[210] C.L. Mendias, J.E. Marcin, D.R. Calerdon, J.A. Faulkner, Contractile properties ofEDL and soleus muscles of myostatin-deficient mice, J. Appl. Physiol. 101 (2006)898–905.

[211] L.A. Whittemore, K. Song, X. Li, J. Aghajanian, M. Davies, S. Girgenrath, J.J. Hill, M.Jalenak, P. Kelley, A. Knight, R. Maylor, D. O'Hara, A. Pearson, A. Quazi, S. Ryerson,X.Y. Tan, K.N. Tomkinson, G.M. Veldman, A. Widom, J.F. Wright, S. Wudyka, L.Zhao, N.M. Wolfman, Inhibition of myostatin in adult mice increases skeletalmuscle mass and strength, Biochem. Biophys. Res. Commun. 300 (2003)965–971.

[212] V. Siriett, M.S. Salerno, C. Berry, G. Nicholas, R. Bower, R. Kambadur, M. Sharma,Antagonism of myostatin enhances muscle regeneration during sarcopenia, Mol.Ther. 15 (2007) 1463–1470.

[213] M.E. Carlson, M. Hsu, I.M. Conboy, Imbalance between pSmad3 and Notchinduces CDK inhibitors in old muscle stem cells, Nature 454 (2008) 528–532.

[214] R.D. Cohn, C. van Erp, J.P. Habashi, A.A. Soleimani, E.C. Klein, M.T. Lisi, M.Gamradt, C.M. ap Rhys, T.M. Holm, B.L. Loeys, F. Ramirez, D.P. Judge, C.W. Ward,H.C. Dietz, Angiotensin II type 1 receptor blockade attenuates TGF-beta-inducedfailure of muscle regeneration in multiple myopathic states, Nat. Med. 13 (2007)204–210.

[215] C.M. Ng, A. Cheng, L.A.Myers, F.Martinez-Murillo, C. Jie, D. Bedja, K.L. Gabrielson, J.M.Hausladen, R.P.Mecham, D.P. Judge, H.C. Dietz, TGF-beta-dependent pathogenesis ofmitral valve prolapse in a mouse model of Marfan syndrome, J. Clin. Invest. 114(2004) 1586–1592.

[216] Z.A. Quinn, C.C. Yang, J.L. Wrana, J.C. McDermott, Smad proteins function asco-modulators for MEF2 transcriptional regulatory proteins, Nucleic Acids Res. 29(2001) 732–742.

[217] D. Liu, J.S. Kang, R. Derynck, TGF-beta-activated Smad3 represses MEF2-dependent transcription in myogenic differentiation, EMBO J. 23 (2004)1557–1566.

[218] M.J. Potthoff, M.A. Arnold, J. McAnally, J.A. Richardson, R. Bassel-Duby, E.N.Olson, Regulation of skeletal muscle sarcomere integrity and postnatal musclefunction by Mef2c, Mol. Cell. Biol. 27 (2007) 8143–8151.

[219] B.A. Reynolds, S. Weiss, Generation of neurons and astrocytes from isolated cells ofthe adult mammalian central nervous system, Science 255 (1992) 1707–1710.


[220] A. Alvarez-Buylla, B. Seri, F. Doetsch, Identification of neural stem cells in theadult vertebrate brain, Brain Res. Bull. 57 (2002) 751–758.

[221] N.D. Bull, P.F. Bartlett, The adult mouse hippocampal progenitor is neurogenicbut not a stem cell, J. Neurosci. 25 (2005) 10815–10821.

[222] F. Doetsch, L. Petreanu, I. Caille, J.M. Garcia-Verdugo, A. Alvarez-Buylla, EGFconverts transit-amplifying neurogenic precursors in the adult brain intomultipotent stem cells, Neuron 36 (2002) 1021–1034.

[223] G.P. Marshall 2nd, E.D. Laywell, T. Zheng, D.A. Steindler, E.W. Scott, Invitro-derived “neural stem cells” function as neural progenitors without thecapacity for self-renewal, Stem Cells 24 (2006) 731–738.

[224] S. Falk, H. Wurdak, L.M. Ittner, F. Ille, G. Sumara, M.T. Schmid, K. Draganova, K.S.Lang, C. Paratore, P. Leveen, U. Suter, S. Karlsson, W. Born, R. Ricci, M. Gotz, L.Sommer, Brain area-specific effect of TGF-beta signaling on Wnt-dependentneural stem cell expansion, Cell Stem Cell 2 (2008) 472–483.

[225] H.L. Chen, D.M. Panchision, Concise review: bone morphogenetic proteinpleiotropism in neural stem cells and their derivatives—alternative pathways,convergent signals, Stem Cells 25 (2007) 63–68.

[226] D.M. Panchision, J.M. Pickel, L. Studer, S.H. Lee, P.A. Turner, T.G. Hazel, R.D.McKay, Sequential actions of BMP receptors control neural precursor cellproduction and fate, Genes Dev. 15 (2001) 2094–2110.

[227] R.E. Gross, M.F. Mehler, P.C. Mabie, Z. Zang, L. Santschi, J.A. Kessler, Bonemorphogenetic proteins promote astroglial lineage commitment by mammaliansubventricular zone progenitor cells, Neuron 17 (1996) 595–606.

[228] W. Li, C.A. Cogswell, J.J. LoTurco, Neuronal differentiation of precursors in theneocortical ventricular zone is triggered by BMP, J. Neurosci. 18 (1998) 8853–8862.

[229] A. Alvarez-Buylla, D.A. Lim, For the long run: maintaining germinal niches in theadult brain, Neuron 41 (2004) 683–686.

[230] F. Doetsch, I. Caille, D.A. Lim, J.M. Garcia-Verdugo, A. Alvarez-Buylla, Subven-tricular zone astrocytes are neural stem cells in the adult mammalian brain, Cell97 (1999) 703–716.

[231] D.A. Lim, A.D. Tramontin, J.M. Trevejo, D.G. Herrera, J.M. Garcia-Verdugo, A.Alvarez-Buylla, Noggin antagonizes BMP signaling to create a niche for adultneurogenesis, Neuron 28 (2000) 713–726.

[232] G.A. Challen, N. Boles, K.K. Lin, M.A. Goodell, Mouse hematopoietic stem cellidentification and analysis, Cytometry A 75 (2009) 14–24.

[233] S.J. Szilvassy, R.K. Humphries, P.M. Lansdorp, A.C. Eaves, C.J. Eaves, Quantitativeassay for totipotent reconstituting hematopoietic stem cells by a competitiverepopulation strategy, Proc. Natl. Acad. Sci. U. S. A. 87 (1990) 8736–8740.

[234] G.J. Spangrude, D.M. Brooks, D.B. Tumas, Long-term repopulation of irradiatedmice with limiting numbers of purified hematopoietic stem cells: in vivoexpansion of stem cell phenotype but not function, Blood 85 (1995) 1006–1016.

[235] B. Dykstra, D. Kent, M. Bowie, L. McCaffrey, M. Hamilton, K. Lyons, S.J. Lee, R.Brinkman, C. Eaves, Long-term propagation of distinct hematopoietic differen-tiation programs in vivo, Cell Stem Cell 1 (2007) 218–229.

[236] C.E. Muller-Sieburg, R.H. Cho, M. Thoman, B. Adkins, H.B. Sieburg, Deterministicregulation of hematopoietic stem cell self-renewal and differentiation, Blood100 (2002) 1302–1309.

[237] C.E. Muller-Sieburg, R.H. Cho, L. Karlsson, J.F. Huang, H.B. Sieburg, Myeloid-biased hematopoietic stem cells have extensive self-renewal capacity butgenerate diminished lymphoid progeny with impaired IL-7 responsiveness,Blood 103 (2004) 4111–4118.

[238] H.B. Sieburg, R.H. Cho, B. Dykstra, N. Uchida, C.J. Eaves, C.E. Muller-Sieburg, Thehematopoietic stem compartment consists of a limited number of discrete stemcell subsets, Blood 107 (2006) 2311–2316.

[239] S. Yamazaki, A. Iwama, S. Takayanagi, K. Eto, H. Ema, H. Nakauchi, TGF-beta as acandidate bone marrow niche signal to induce hematopoietic stem cellhibernation, Blood 113 (2009) 1250–1256.

[240] S. Yamazaki, H. Ema, G. Karlsson, T. Yamaguchi, H. Miyoshi, S. Shioda, M.M.Taketo, S. Karlsson, A. Iwama, H. Nakauchi, Nonmyelinating Schwann cellsmaintain hematopoietic stem cell hibernation in the bone marrow niche, Cell147 (2011) 1146–1158.

[241] J. Larsson, U. Blank, H. Helgadottir, J.M. Bjornsson, M. Ehinger, M.J. Goumans, X.Fan, P. Leveen, S. Karlsson, TGF-beta signaling-deficient hematopoietic stemcells have normal self-renewal and regenerative ability in vivo despite increasedproliferative capacity in vitro, Blood 102 (2003) 3129–3135.

[242] J. Larsson, U. Blank, J. Klintman, M. Magnusson, S. Karlsson, Quiescence ofhematopoietic stem cells and maintenance of the stem cell pool is notdependent on TGF-beta signaling in vivo, Exp. Hematol. 33 (2005) 592–596.

[243] V.P. Kale, A.A. Vaidya, Molecular mechanisms behind the dose-dependentdifferential activation of MAPK pathways induced by transforming growthfactor-beta1 in hematopoietic cells, Stem Cells Dev. 13 (2004) 536–547.

[244] G.A. Challen, N.C. Boles, S.M. Chambers, M.A. Goodell, Distinct hematopoieticstem cell subtypes are differentially regulated by TGF-beta1, Cell Stem Cell 6(2010) 265–278.

[245] L.E. Ailles, I.L. Weissman, Cancer stem cells in solid tumors, Curr. Opin.Biotechnol. 18 (2007) 460–466.

[246] C. Scheel, R.A. Weinberg, Phenotypic plasticity and epithelial–mesenchymaltransitions in cancer and normal stem cells? Int. J. Cancer 129 (2011)2310–2314.

[247] K. Polyak, R.A. Weinberg, Transitions between epithelial and mesenchymal states:acquisition of malignant and stem cell traits, Nat. Rev. Cancer 9 (2009) 265–273.

[248] M. Al-Hajj, M.S. Wicha, A. Benito-Hernandez, S.J. Morrison, M.F. Clarke,Prospective identification of tumorigenic breast cancer cells, Proc. Natl. Acad.Sci. U. S. A. 100 (2003) 3983–3988.

[249] C. Ginestier, M.H. Hur, E. Charafe-Jauffret, F. Monville, J. Dutcher, M. Brown, J.Jacquemier, P. Viens, C.G. Kleer, S. Liu, A. Schott, D. Hayes, D. Birnbaum, M.S.

Wicha, G. Dontu, ALDH1 is a marker of normal and malignant human mammarystem cells and a predictor of poor clinical outcome, Cell Stem Cell 1 (2007)555–567.

[250] A.P. Morel, M. Lievre, C. Thomas, G. Hinkal, S. Ansieau, A. Puisieux, Generation ofbreast cancer stem cells through epithelial–mesenchymal transition, PLoS One 3(2008) e2888.

[251] J.T. Buijs, G. van der Horst, C. van den Hoogen, H. Cheung, B. de Rooij, J. Kroon, M.Petersen, P.G. van Overveld, R.C. Pelger, G. van der Pluijm, The BMP2/7heterodimer inhibits the human breast cancer stem cell subpopulation andbone metastases formation, Oncogene 31 (2012) 2164–2174.

[252] B. Tang, N. Yoo, M. Vu, M. Mamura, J.S. Nam, A. Ooshima, Z. Du, P.Y. Desprez, M.R.Anver, A.M. Michalowska, J. Shih, W.T. Parks, L.M. Wakefield, Transforminggrowth factor-beta can suppress tumorigenesis through effects on the putativecancer stem or early progenitor cell and committed progeny in a breast cancerxenograft model, Cancer Res. 67 (2007) 8643–8652.

[253] P.C. Hermann, S.L. Huber, T. Herrler, A. Aicher, J.W. Ellwart, M. Guba, C.J. Bruns, C.Heeschen, Distinct populations of cancer stem cells determine tumor growthand metastatic activity in human pancreatic cancer, Cell Stem Cell 1 (2007)313–323.

[254] C. Li, D.G. Heidt, P. Dalerba, C.F. Burant, L. Zhang, V. Adsay, M. Wicha, M.F. Clarke,D.M. Simeone, Identification of pancreatic cancer stem cells, Cancer Res. 67(2007) 1030–1037.

[255] E. Lonardo, P.C. Hermann, M.T. Mueller, S. Huber, A. Balic, I. Miranda-Lorenzo, S.Zagorac, S. Alcala, I. Rodriguez-Arabaolaza, J.C. Ramirez, R. Torres-Ruiz, E. Garcia,M. Hidalgo, D.A. Cebrian, R. Heuchel, M. Lohr, F. Berger, P. Bartenstein, A. Aicher,C. Heeschen, Nodal/Activin signaling drives self-renewal and tumorigenicity ofpancreatic cancer stem cells and provides a target for combined drug therapy,Cell Stem Cell 9 (2011) 433–446.

[256] L. Ricci-Vitiani, D.G. Lombardi, E. Pilozzi, M. Biffoni, M. Todaro, C. Peschle, R. DeMaria, Identification and expansion of human colon-cancer-initiating cells,Nature 445 (2007) 111–115.

[257] C.A. O'Brien, A. Pollett, S. Gallinger, J.E. Dick, A human colon cancer cell capableof initiating tumour growth in immunodeficient mice, Nature 445 (2007)106–110.

[258] Y. Lombardo, A. Scopelliti, P. Cammareri, M. Todaro, F. Iovino, L. Ricci-Vitiani, G.Gulotta, F. Dieli, R. de Maria, G. Stassi, Bone morphogenetic protein 4 inducesdifferentiation of colorectal cancer stem cells and increases their response tochemotherapy in mice, Gastroenterology 140 (2011) 297–309.

[259] L. Strizzi, K.M. Hardy, G.T. Kirsammer, P. Gerami, M.J. Hendrix, Embryonicsignaling in melanoma: potential for diagnosis and therapy, Lab. Invest. 91(2011) 819–824.

[260] D. Fang, T.K. Nguyen, K. Leishear, R. Finko, A.N. Kulp, S. Hotz, P.A. Van Belle, X.Xu, D.E. Elder, M. Herlyn, A tumorigenic subpopulation with stem cell propertiesin melanomas, Cancer Res. 65 (2005) 9328–9337.

[261] E. Monzani, F. Facchetti, E. Galmozzi, E. Corsini, A. Benetti, C. Cavazzin, A. Gritti,A. Piccinini, D. Porro, M. Santinami, G. Invernici, E. Parati, G. Alessandri, C.A. LaPorta, Melanoma contains CD133 and ABCG2 positive cells with enhancedtumourigenic potential, Eur. J. Cancer 43 (2007) 935–946.

[262] E. Quintana, M. Shackleton, H.R. Foster, D.R. Fullen, M.S. Sabel, T.M. Johnson, S.J.Morrison, Phenotypic heterogeneity among tumorigenic melanoma cells frompatients that is reversible and not hierarchically organized, Cancer Cell 18(2010) 510–523.

[263] A. Roesch, M. Fukunaga-Kalabis, E.C. Schmidt, S.E. Zabierowski, P.A. Brafford, A.Vultur, D. Basu, P. Gimotty, T. Vogt, M. Herlyn, A temporarily distinctsubpopulation of slow-cycling melanoma cells is required for continuoustumor growth, Cell 141 (2010) 583–594.

[264] K.S. Hoek, C.R. Goding, Cancer stem cells versus phenotype-switching inmelanoma, Pigment Cell Melanoma Res. 23 (2010) 746–759.

[265] J.M. Topczewska, L.M. Postovit, N.V. Margaryan, A. Sam, A.R. Hess, W.W.Wheaton, B.J. Nickoloff, J. Topczewski, M.J. Hendrix, Embryonic and tumorigenicpathways converge via Nodal signaling: role in melanoma aggressiveness, Nat.Med. 12 (2006) 925–932.

[266] L.M. Postovit, N.V. Margaryan, E.A. Seftor, D.A. Kirschmann, A. Lipavsky, W.W.Wheaton, D.E. Abbott, R.E. Seftor, M.J. Hendrix, Human embryonic stem cellmicroenvironment suppresses the tumorigenic phenotype of aggressive cancercells, Proc. Natl. Acad. Sci. U. S. A. 105 (2008) 4329–4334.

[267] L. Strizzi, K.M. Hardy, E.A. Seftor, F.F. Costa, D.A. Kirschmann, R.E. Seftor, L.M.Postovit, M.J. Hendrix, Development and cancer: at the crossroads of Nodal andNotch signaling, Cancer Res. 69 (2009) 7131–7134.

[268] M. Schober, E. Fuchs, Tumor-initiating stem cells of squamous cell carcinomasand their control by TGF-beta and integrin/focal adhesion kinase (FAK)signaling, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 10544–10549.

[269] S. Ehata, E. Johansson, R. Katayama, S. Koike, A. Watanabe, Y. Hoshino, Y.Katsuno, A. Komuro, D. Koinuma, M.R. Kano, M. Yashiro, K. Hirakawa, H.Aburatani, N. Fujita, K. Miyazono, Transforming growth factor-beta decreasesthe cancer-initiating cell population within diffuse-type gastric carcinoma cells,Oncogene 30 (2011) 1693–1705.

[270] Y. Katsuno, S. Ehata, M. Yashiro, K. Yanagihara, K. Hirakawa, K. Miyazono,Coordinated expression of REG4 and aldehyde dehydrogenase 1 regulatingtumourigenic capacity of diffuse-type gastric carcinoma-initiating cells isinhibited by TGF-beta, J. Pathol. (2012), http://dx.doi.org/10.1002/path.4020.

[271] L. Cao, M. Shao, J. Schilder, T. Guise, K.S. Mohammad, D. Matei, Tissuetransglutaminase links TGF-beta, epithelial to mesenchymal transition and astem cell phenotype in ovarian cancer, Oncogene 31 (2012) 2521–2534.

[272] A. Kobayashi, H. Okuda, F. Xing, P.R. Pandey, M. Watabe, S. Hirota, S.K. Pai, W.Liu, K. Fukuda, C. Chambers, A. Wilber, K. Watabe, Bone morphogenetic protein

http://dx.doi.org/10.1002/path.4020


7 in dormancy and metastasis of prostate cancer stem-like cells in bone, J. Exp.Med. 208 (2011) 2641–2655.

[273] S.K. Singh, C. Hawkins, I.D. Clarke, J.A. Squire, J. Bayani, T. Hide, R.M. Henkelman,M.D. Cusimano, P.B. Dirks, Identification of human brain tumour initiating cells,Nature 432 (2004) 396–401.

[274] S.G. Piccirillo, B.A. Reynolds, N. Zanetti, G. Lamorte, E. Binda, G. Broggi, H. Brem, A.Olivi, F. Dimeco, A.L. Vescovi, Bonemorphogenetic proteins inhibit the tumorigenicpotential of human brain tumour-initiating cells, Nature 444 (2006) 761–765.

[275] J. Lee, M.J. Son, K. Woolard, N.M. Donin, A. Li, C.H. Cheng, S. Kotliarova, Y. Kotliarov, J.Walling, S. Ahn, M. Kim, M. Totonchy, T. Cusack, C. Ene, H. Ma, Q. Su, J.C. Zenklusen,W. Zhang, D. Maric, H.A. Fine, Epigenetic-mediated dysfunction of the bonemorphogenetic protein pathway inhibits differentiation of glioblastoma-initiatingcells, Cancer Cell 13 (2008) 69–80.

[276] S. Penuelas, J. Anido, R.M. Prieto-Sanchez, G. Folch, I. Barba, I. Cuartas, D.Garcia-Dorado, M.A. Poca, J. Sahuquillo, J. Baselga, J. Seoane, TGF-beta increasesglioma-initiating cell self-renewal through the induction of LIF in humanglioblastoma, Cancer Cell 15 (2009) 315–327.

[277] H. Ikushima, T. Todo, Y. Ino, M. Takahashi, K. Miyazawa, K. Miyazono, AutocrineTGF-beta signaling maintains tumorigenicity of glioma-initiating cells throughSry-related HMG-box factors, Cell Stem Cell 5 (2009) 504–514.

[278] M. Zhang, S. Kleber, M. Rohrich, C. Timke, N. Han, J. Tuettenberg, A. Martin-Villalba,J. Debus, P. Peschke, U. Wirkner, M. Lahn, P.E. Huber, Blockade of TGF-betasignaling by the TGFbetaR-I kinase inhibitor LY2109761 enhances radiationresponse and prolongs survival in glioblastoma, Cancer Res. 71 (2011) 7155–7167.

[279] D. Bonnet, J.E. Dick, Human acute myeloid leukemia is organized as a hierarchythat originates from a primitive hematopoietic cell, Nat. Med. 3 (1997) 730–737.

[280] L. Jin, K.J. Hope, Q. Zhai, F. Smadja-Joffe, J.E. Dick, Targeting of CD44 eradicates humanacute myeloid leukemic stem cells, Nat. Med. 12 (2006) 1167–1174.

[281] D.S. Krause, K. Lazarides, U.H. von Andrian, R.A. Van Etten, Requirement forCD44 in homing and engraftment of BCR-ABL-expressing leukemic stem cells,Nat. Med. 12 (2006) 1175–1180.

[282] K. Naka, T. Hoshii, T. Muraguchi, Y. Tadokoro, T. Ooshio, Y. Kondo, S. Nakao, N.Motoyama, A. Hirao, TGF-beta–FOXO signalling maintains leukaemia-initiatingcells in chronic myeloid leukaemia, Nature 463 (2010) 676–680.

[283] R. Quere, G. Karlsson, F. Hertwig,M. Rissler, B. Lindqvist, T. Fioretos, P. Vandenberghe,M.L. Slovak, J. Cammenga, S. Karlsson, Smad4 binds Hoxa9 in the cytoplasm andprotects primitive hematopoietic cells against nuclear activation by Hoxa9 andleukemia transformation, Blood 117 (2011) 5918–5930.

[284] X. Guo, H. Oshima, T. Kitmura, M.M. Taketo, M. Oshima, Stromal fibroblastsactivated by tumor cells promote angiogenesis in mouse gastric cancer, J. Biol.Chem. 283 (2008) 19864–19871.

[285] E.M. Zeisberg, S. Potenta, L. Xie, M. Zeisberg, R. Kalluri, Discovery of endothelialto mesenchymal transition as a source for carcinoma-associated fibroblasts,Cancer Res. 67 (2007) 10123–10128.

[286] M. Quante, S.P. Tu, H. Tomita, T. Gonda, S.S. Wang, S. Takashi, G.H. Baik, W.Shibata, B. Diprete, K.S. Betz, R. Friedman, A. Varro, B. Tycko, T.C. Wang, Bonemarrow-derived myofibroblasts contribute to the mesenchymal stem cell nicheand promote tumor growth, Cancer Cell 19 (2011) 257–272.

[287] K. McLean, Y. Gong, Y. Choi, N. Deng, K. Yang, S. Bai, L. Cabrera, E. Keller, L.McCauley, K.R. Cho, R.J. Buckanovich, Human ovarian carcinoma-associatedmesenchymal stem cells regulate cancer stem cells and tumorigenesis viaaltered BMP production, J. Clin. Invest. 121 (2011) 3206–3219.

TGF-β family signaling in stem cells

Documents

Transcript of TGF-β family signaling in stem cells