Update TRENDS in Neurosciences Vol.28 No.2 February 200562

3 Kang, J. et al. (1998) Astrocyte-mediated potentiation of inhibitorysynaptic transmission. Nat. Neurosci. 1, 683–692

4 Araque, A. et al. (2002) Synaptically released acetylcholine evokesCa2C elevations in astrocytes in hippocampal slices. J. Neurosci. 22,2443–2450

5 Lin, S. and Bergles, D.E. (2004) Synaptic signaling betweenGABAergic interneurons and oligodendrocyte precursor cells in thehippocampus. Nat. Neurosci. 7, 24–32

6 Bergles, D.E. et al. (2000) Glutamatergic synapses on oligodendrocyteprecursor cells in the hippocampus. Nature 405, 187–191

7 Mucrick-Donnon, L.A. et al. (1993) Postsynaptic potentials mediatedby GABA and dopamine evoked in stellate glial cells of the pituitarypars intermedia. J. Neurosci. 13, 4660–4668

8 Dawson, M.R. et al. (2003) NG2-expressing glial progenitor cells: anabundant and widespread population of cycling cells in the adult ratCNS. Mol. Cell. Neurosci. 24, 476–488

9 Lin, S. and Bergles, D.E. (2002) Physiological characteristics of NG2-expressing glial cells. J. Neurocytol. 31, 537–549

10 Barres, B.A. et al. (1990) Ion channel expression by white matter glia:the O-2A glia progenitor cell. Neuron 4, 507–524

11 Patneau, D.K. et al. (1994) Glial cells of the oligodendrocyte lineageexpress both kainate- and AMPA-preferring subtypes of glutamatereceptor. Neuron 12, 357–371

12 Von Blankenfeld, G. et al. (1991) Expression and developmentalregulation of a GABAA receptor in cultured murine cells of theoligodendrocyte lineage. Eur. J. Neurosci. 3, 310–316

13 Matsui, K. and Jahr, C.E. (2003) Ectopic release of synaptic vesicles.Neuron 40, 1173–1183

14 Stegmuller, J. et al. (2003) The proteoglycan NG2 is complexed with

Corresponding author: Simeone, A. (antonio.simeone@kcl.ac.uk).Available online 30 November 2004

www.sciencedirect.com

a-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) recep-tors by the PDZ glutamate receptor interaction protein (GRIP) in glialprogenitor cells. J. Biol. Chem. 278, 3590–3598

15 Gallo, V. et al. (1996) Oligodendrocyte progenitor cell proliferation andlineage progression are regulated by glutamate receptor-mediated KC

channel block. J. Neurosci. 16, 2659–267016 Liu, H.N. and Almazan, G. (1995) Glutamate induces c-fos proto-

oncogene expression and inhibits proliferation in oligodendrocyteprogenitors. Eur. J. Neurosci. 7, 2355–2363

17 Yuan, X. et al. (1998) A role for glutamate and its receptors in theregulation of oligodendrocyte development in cerebellar tissue slices.Development 125, 2901–2914

18 Owens, D.F. and Kriegstein, A.R. (2002) Is there more to GABA thansynaptic inhibition? Nat. Rev. Neurosci. 3, 715–727

19 Dawson, M.L. et al. (2000) NG2-expressing cells in the central nervoussystem: are they oligodendroglial progenitors? J. Neurosci. Res. 61,471–479

20 Horner, P.J. et al. (2002) Defining the NG2-expressing cell of the adultCNS. J. Neurocytol. 31, 469–480

21 Goldman, S. (2003) Glia as neural progenitor cells. Trends Neurosci.26, 596

22 Deng, W. et al. (2003) Calcium-permeable AMPA/kainite receptorsmediate toxicity and precondition by oxygen-glucose deprivation inoligodendrocyte precursors.Proc.Natl.Acad.Sci.U.S.A.100, 6801–6806

23 Follett, P.L. et al. (2004) Glutamate receptor-mediated oligodendro-cyte toxicity in periventricular leukomalacia: a protective role fortopiramate. J. Neurosci. 24, 4412–4420

0166-2236/$ - see front matter Q 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.tins.2004.11.010

Genetic control of dopaminergic neuron differentiation

Antonio Simeone

MRC Centre for Developmental Neurobiology, 4th Floor, New Hunt’s House, Guy’s Campus, King’s College London, London Bridge,

London SE1 1UL, UK

The mesencephalic dopaminergic (mesDA) system con-

trols movement and emotional behaviour, and its

degeneration causes Parkinson’s disease and other

psychiatric disorders. Recent findings are leading to

better understanding of the genetic control of gener-

ation and functioning of the mesDA system. This

advancement could disclose new perspectives for

therapeutic approaches of mesDA-related disorders.

Introduction

Development of the vertebrate CNS is a multi-stepprocess. It involves subdivision of the anterior neuroecto-derm into broad territories corresponding to the forebrain,midbrain and hindbrain and, subsequently, generation ofspecific neural cell-types within each of these regions[1–3]. These events are controlled by inducing signals anddepend on the responding ability of target cells. Theisthmic organizer at the midbrain–hindbrain boundary

and the floor plate and roof plate of the neural tube areinducing centres operating along the anterior–posteriorand dorsal–ventral axes [1,4]. The intersection of signalsemitted from these centres provides positional infor-mation to neuronal precursors which, in turn, arecompetent to interpret these signals and convert theirinstructions into specific fates.

Signaling molecules encoded by Shh and Fgf8 haveprovided important results on the mechanism controllingidentity, growth and survival of neuronal precursors[1,5–7]. In the rostral CNS, transcription factors belong-ing to the Otx, En, Pax, Gbx, Dlx and Emx families arerequired to specify regional identities and to controlneuronal differentiation. A crucial task is, therefore, toelucidate the network of interactions and the hierarchy ofmolecular events that gradually confer full differentiationto a specific type of neuronal progenitor.

Although it is widely accepted that sonic hedgehog (Shh)and fibroblast growth factor (FGF)8 signaling moleculesinstruct identity and position of dopaminergic neuronprogenitors [7,8], the subsequent steps differentiation are

TRENDS in Neurosciences

E12

E14

SN

SN

E18

Pitx3 TH Pitx3 +TH

SN

Late survival ofmesDA neurons

Wild-type

VTA

ak/ak(Pitx3 null) Lmx1b–/– Nr4a2–/–

Specificationof mesDA neurons

Neurotransmitterphenotype

VTA

VTA

Figure 1. Abnormalities of the mesDA system in ak/ak, Lmx1bK/K and Nr4a2K/K

mutant mice at mid-gestation (E12 and E14) and late-gestation (E18). Inak/akmutants,

some VTA and SN neurons were selectively lost at E12; in Lmx1bK/K mutants, Pitx3

was not expressed in TH-positive neurons at E12; and in Nr4a2K/K mutants,

TH expression was not activated in Pitx3-positive neurons. These findings

suggest that two major pathways converge towards terminal differentiation of

mesDA neurons: one for neurotransmitter phenotype requiring Nr4a2, and the

other for specification of dopaminergic neurons requiring Lmx1b and Pitx3. In

this context, Nr4a2 is also required for late survival of mesDA neurons.

Therefore, Pitx3 and Lmx1b are required for generation of post-mitotic mesDA

neurons, whereas Nr4a2 is necessary to provide mesDA neurons with a crucial

phenotypic feature. The red crosses indicate which region (SN, VTA or both)

mesDA neurons were lost.

Update TRENDS in Neurosciences Vol.28 No.2 February 2005 63

not yet fully understood. Recent findings are unmaskingthis process, and in this context the role ofPitx3 (also knownas Ptx3) is of particular relevance [9–12].

Pitx3 controls development of dopaminergic neurons of

the substantia nigra

Neurons of the mesencephalic dopaminergic (mesDA)system are allocated in three different major subgroupscorresponding to the ventral tegmental area (VTA), thesubstantia nigra (SN) and the retrorubral field (RRF), andthey project to the forebrain with a specific pattern ofinnervations. Dopaminergic neurons of the VTA project tothe ventromedial striatum, nucleus accumbens, temporallobe and olfactory tubercle; those of the SN innervate thedorsolateral striatum, and so do those of the RRF [13].

The clinical relevance of mesDA neurons is highbecause impairment in their survival and/or developmentis responsible for abnormal control of voluntary move-ment, behaviour and cognition [14,15]. Indeed, degener-ation of mesDA neurons, particularly those of the SN,underlies Parkinson’s disease, and abnormal dopamin-ergic signaling is involved in psychiatric disorders(e.g. schizophrenia) and addictive behavioural disorders.

The Pitx3 gene contains a bicoid-related homeobox andbelongs to the Pitx family of transcription factor genes. Inthe brain, Pitx3 is expressed in the mesDA neurons fromembryonic day (E)11.0 and throughout adult life; itsexpression is, therefore, concomitant with the transcrip-tional activation of the tyrosine hydroxylase gene (TH)[16] (Figure 1a). Recently, four papers [9–12] havereported important results on the development of mesDAneurons in the aphakia (ak) homozygous mouse model,which lacks detectable level of Pitx3 transcripts and canbe considered as a Pitx3-null mutant [17]. Remarkably, inak mutants, TH-positive neurons of the SN were lost butthose of the VTA were less affected, suggesting that Pitx3is required for formation of the SN and the specificationand/or survival of this subpopulation of mesDA neurons(Figure 1). This result implies that differentiation ofrestricted mesDA neuron subpopulations could be underthe control of different factors, or that different subpopu-lations differentially respond to the same factor. In thiscontext, Pitx3 expression has been studied in TH-positiveneurons of the SN and VTA. This analysis has providedcontrasting results: van den Munckhof and colleagues [9]reported that Pitx3 was expressed in w50% of theTH-positive neurons of the VTA and in most of theTH-positive neurons in the ventral tier of the SN, whereasSmidt and colleagues [12] showed that Pitx3 wasexpressed in all the TH-positive neurons of the VTA andSN. This apparent difference has yet to be resolved.

In ak mutant mice, loss of SN neurons also resulted in asevere depletion of TH-staining from the SN and loss ofdopaminergic projections from the SN to the dorsolateralstriatum. In line with the less severe effects of the akmutation on the VTA, TH staining in this area anddopaminergic projections from it to the ventromedialstriatum, nucleus accumbens and olfactory tubercle wereunaffected [9,12]. Behavioural tests of the locomotor activityof ak mice have shown an overall reduced activity [9,12].Therefore, Pitx3 is crucial for the development of the SN

www.sciencedirect.com

subpopulation of mesDA neurons (Figure 1). These ana-tomical defects, together with locomotor hypoactivity,indicate some similarity with the Parkinson’s diseasephenotype, even though no data show a casual link betweenhuman Parkinson’s disease and Pitx3 mutations. Indeed,two Pitx3 mutations have been identified in families withautosomal dominant cataracts and, although this pheno-type closely matches the eye defect observed in ak mutantmice, no Parkinsonian symptoms have been described inthese patients [18].

Genetic cascade controlling late differentiation and

survival of the mesDA system

Two additional transcription factor genes, Nr4a2 (alsoknown as Nurr1) and Lmx1b, are involved in specificationof the mesDA neurons. Nr4a2 is a member of the nuclearreceptor superfamily of transcription factors and isexpressed in the ventral midbrain, as well as in otherregions of the developing and adult brain.Nr4a2 expressionis detectable at E10.5 in immature mesDA neuronprecursors and, therefore, precedes TH expression andemergence of differentiated dopaminergic neurons [19].Nr4a2-null mice died soon after birth and lacked dopamin-ergic neurons (Figure 1). Analysis of Nr4a2-null mutantembryos revealed that TH expression could not be detectedat early embryonic stages, whereas Pitx3 was expressednormally until late gestation, when dopaminergic neuronsdied, possibly because of increased apoptosis [19,20].

Update TRENDS in Neurosciences Vol.28 No.2 February 200564

Therefore, although it is unclear whether Nr4a2 canpromote survival of dopaminergic neurons through apositive effect on expression of neurotrophic factors, itseems to be required for proper neurotransmitterexpression. Moreover, the dopamine transporter and thevesicular monoamine transporter-2 depend on Nr4a2activity but not on Pitx3 [21]. These data, and the findingthat inakmice residual TH-positive neurons exhibit normalexpression of Nr4a2, strongly suggest that Pitx3 and Nr4a2control different aspects of mesDA neuron differentiation.

Lmx1b, a Lim homeodomain transcription factor, isexpressed in VTA and SN precursors before the onset ofNr4a2 and Pitx3 expression [22]. Limx1b-null mice lackeddopaminergic neurons at late gestation and, interestingly,failed to activate the expression of Pitx3, whereas theinitial expression of TH and Nr4a2 were unaffected [22](Figure 1). This finding, together with those indicatingthat expression of Lmx1b is normal in TH-positive cells ofak and Nr4a2 mutants, suggest that Lmx1b is involved inactivation of Pitx3 in mesDA neurons, whereas Nr4a2 isinvolved in terminal differentiation by controlling neuro-transmitter expression (Figure 2). This suggests thatterminal differentiation of the mesDA system involves twoseparate pathways that converge to provide the functionalfeatures of a differentiated mesDA neuron. Moreover,because Pitx3 is required mainly for specification of theSN, it is tempting to speculate that Lmx1b activates Pitx3expression in the SN but that Lmx1b or another factoractivates expression of a VTA-specific differentiation

TH

Nr4

a2

Spe

Induction of dopaminergicprogenitors

Conversion of inducing signainto mesDA progenitor cell

identity

Proliferation of dopaminergicneuronal progenitors

Ventrolateral migration

Terminaldifferentiation

Neurotransmitterphenotype

Figure 2. Developmental steps and genetic cascade controlling induction and fate of me

and FGF8, which is converted into dopaminergic progenitor cell identity. In this proces

expression domains, and Otx2 in the ventral midbrain suppresses hindbrain fate and

modulate proliferation of neuronal progenitors before their ventrolateral migration and

SN and VTA sub-populations. Neurotransmitter phenotype (i.e. activation of TH) and la

requires Pitx3; and both Pitx3 activation and early maintenance of SN and VTA neurons

specification factor equivalent to Pitx3 in the SN has yet to be identified.

www.sciencedirect.com

factor in the VTA. The complexity of the mesDA systemis already known to require other factors; for example,engrailed genes (En1 and En2) are involved in itsmaintenance and survival [23].

Early specification of mesDA precursors

The early commitment of neuronal precursors of the ventralmesencephalon depends on the inducing properties of FGF8and Shh signals emitted at the isthmic organizer and floorplate, respectively [7,8] (Figure 1b). Shh and FGF8, incooperation with FGF4, are also required for the inductionand positioning of cells containing 5-hydroxytryptamine(5-HT or serotonin) in the rostral hindbrain [8].

Recently, it has been reported that Otx1 and Otx2,through a dose-dependent repressive effect, coordinateproper positioning of Shh and Fgf8 expression [24].Indeed, reduced levels of Otx proteins in proximity of thealar–basal plate boundary of the midbrain and at themesencephalic side of the midbrain–hindbrain boundaryresulted in dorsal shift of Shh expression and anteriorexpansion of the Fgf8 domain (Figure 2). This in turnaffected the identity, fate and proliferation of neuronalprogenitors of the ventral midbrain and resulted in aremarkable increase in numbers of mesDA neurons. Otx2also plays a more direct role in differentiation ofprogenitors in the ventral mesencephalon: selective abla-tion of Otx2 in this region was reflected in a drastic changein the identity code exhibited by neuronal progenitors, toone similar to that of the ventral rostral hindbrain, even

TRENDS in Neurosciences

Supression of hindbrain fate

SN Pitx3

Lmx1

b

cification

ShhFGF8

Positioning control of Shhand FGF8 expression

Otx2antagonism

Shh(FGF8?)

ls

VTA

sDA progenitors. Induction of mesDA neurons requires the combined effect of Shh

s, Otx proteins control in a dose-dependent manner the position of Shh and FGF8

differentiation of neurons containing 5-HT. Shh (and perhaps FGF8) is required to

terminal differentiation in terms of neurotransmitter phenotype and specification of

te survival of mesDA neurons are controlled by Nr4a2; specification of SN neurons

require Lmx1b. For the VTA subpopulation of mesDA neurons (broken bracket), a

Update TRENDS in Neurosciences Vol.28 No.2 February 2005 65

though these neuronal precursors were rostral to theregion of Fgf8 expression. This resulted in a strikingreduction in numbers of mesDA neurons and generation ofan equivalent number of 5-HT-containing neurons [25](Figure 2). These data indicate that Otx2 is required toprovide midbrain neuronal precursors with a specificdifferentiation code, suppressing that of the rostralhindbrain by conferring competence for interpreting Shhand FGF8 signaling activities.

Conclusions and perspectives

Mouse models here discussed are converging toelucidate genetic control of differentiation of themesDA system. In this context a single gene, Pitx3,is required for the terminal differentiation of the SNsubpopulation of mesDA neurons. Importantly, the akmutant model exhibits phenotypic features in commonwith human Parkinson’s disease, and it could be apowerful tool for studying molecular and behaviouralaspects of the mesDA system, as well as for generatingnovel pharmacological approaches. Such knowledgemight also contribute to the generation of geneticallymodified stem cells capable of directing generation ofdopaminergic neurons equivalent to those of the SN.

References

1 Jessell, T.M. (2000) Neuronal specification in the spinal cord:inductive signals and transcriptional codes. Nat. Rev. Genet. 1, 20–29

2 Rubenstein, J.L. et al. (1998) Regionalization of the prosencephalicneural plate. Annu Rev Neurosci. 21, 445–477

3 Lumsden, A. and Krumlauf, R. (1996) Patterning the vertebrateneuraxis. Science 274, 1109–1115

4 Wurst, W. and Bally-Cuif, L. (2001) Neural plate patterning:upstream and downstream of the isthmic organizer. Nat. Rev.Neurosci. 2, 99–108

5 Crossley, P.H. et al. (1996) Midbrain development induced by FGF8 inthe chick embryo. Nature 380, 66–68

6 Briscoe, J. and Ericson, J. (2001) Specification of neuronal fates in theventral neural tube. Curr. Opin. Neurobiol. 11, 43–49

7 Hynes, M. and Rosenthal, A. (1999) Specification of dopaminergic andserotonergic neurons in the vertebrate CNS. Curr. Opin. Neurobiol. 9,26–36

8 Ye, W. et al. (1998) FGF and Shh signals control dopaminergic andserotonergic cell fate in the anterior neural plate. Cell 93, 755–766

Corresponding author: Burbach, J.P.H. (j.p.h.burbach@med.uu.nl).Available online 30 November 2004

www.sciencedirect.com

9 van den Munckhof, P. et al. (2003) Pitx3 is required for motor activityand for survival of a subset of midbrain dopaminergic neurons.Development 130, 2535–2542

10 Nunes, I. et al. (2003) Pitx3 is required for development ofsubstantia nigra dopaminergic neurons. Proc. Natl. Acad. Sci.U. S. A. 100, 4245–4250

11 Hwang, D.Y. et al. (2003) Selective loss of dopaminergic neurons in thesubstantia nigra of Pitx3-deficient aphakia mice. Brain Res. Mol.Brain Res. 114, 123–131

12 Smidt, M.P. et al. (2004) Early developmental failure of substantianigra dopamine neurons in mice lacking the homeodomain gene Pitx3.Development 131, 1145–1155

13 Bjorklund, A. and Lindvall, O. (1984) Dopamine containing systems inthe CNS. In Handbook of Chemical Neuroanatomy (Bjorklund, A. andHokfelt, T. eds), pp. 55–122, Elsevier

14 Egan, M.F. and Weinberger, D.R. (1997) Neurobiology of schizo-phrenia. Curr. Opin. Neurobiol. 7, 701–707

15 Jellinger, K.A. (2001) The pathology of Parkinson’s disease. InParkinson’s Diseases: Advances in Neurology (Vol. 86) (Calne, D.and Calne, S. eds), pp. 52–72, Lippincott, Williams and Wilkins

16 Smidt, M.P. et al. (1997) A homeodomain gene Ptx3 has highlyrestricted brain expression in mesencephalic dopaminergic neurons.Proc. Natl. Acad. Sci. U. S. A. 94, 13305–13310

17 Rieger, D.K. et al. (2001) A double-deletion mutation in the Pitx3 genecauses arrested lens development in aphakia mice. Genomics 72, 61–72

18 Semina, E.V. et al. (1998) A novel homeobox gene PITX3 is mutated infamilies with autosomal-dominant cataracts and ASMD. Nat. Genet.19, 167–170

19 Saucedo-Cardenas, O. et al. (1998) Nurr1 is essential for the inductionof the dopaminergic phenotype and the survival of ventral mesence-phalic late dopaminergic precursor neurons. Proc. Natl. Acad. Sci.U. S. A. 95, 4013–4018

20 Zetterstrom, R.H. et al. (1997) Dopamine neuron agenesis in Nurr1-deficient mice. Science 276, 248–250

21 Smits, S.M. et al. (2003) Involvement of Nurr1 in specifying theneurotransmitter identity of ventral midbrain dopaminergic neurons.Eur. J. Neurosci. 18, 1731–1738

22 Smidt, M.P. et al. (2000) A second independent pathway fordevelopment of mesencephalic dopaminergic neurons requiresLmx1b. Nat. Neurosci. 3, 337–341

23 Simon, H.H. et al. (2001) Fate of midbrain dopaminergic neuronscontrolled by the engrailed genes. J. Neurosci. 21, 3126–3134

24 Puelles, E. et al. (2003) Otx dose-dependent integrated control ofantero-posterior and dorso-ventral patterning of midbrain. Nat.Neurosci. 6, 453–460

25 Puelles, E. et al. (2004) Otx2 regulates the extent, identity and fate ofneuronal progenitor domains in the ventral midbrain. Development131, 2037–2048

0166-2236/$ - see front matter Q 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.tins.2004.11.007

Research Focus Response

Response to Simeone: Coexpression of Pitx3 withtyrosine hydroxylase in midbrain dopaminergicneuronsJ. Peter H. Burbach and Marten P. Smidt

Rudolf Magnus Institute of Neuroscience, Department of Pharmacology and Anatomy, University Medical Center Utrecht,

Universiteitsweg 100, 3584 CG Utrecht, The Netherlands

In his Research Focus article [1], Antonio Simeonehighlights the distinct role of the homeobox transcriptionfactor Pitx3 in terminal specification of midbrain dopa-minergic neurons, as revealed in the loss of dopaminergic