bone morphogenetic protein-2 promotes survival and differentiation of striatal gabaergic neurons in...

Bone Morphogenetic Protein-2 Promotes Survival andDifferentiation of Striatal GABAergic Neurons in

the Absence of Glial Cell Proliferation

*†Akira Hattori, ‡Masahiro Katayama, *§Shoji Iwasaki, ‡Kazuhiro Ishii, †Masafumi Tsujimoto,and *§Michiaki Kohno

*Laboratory of Cell Biology, Gifu Pharmaceutical University, Gifu;†Laboratory of Cellular Biochemistry, Institute of Physicaland Chemical Research (RIKEN), Wako, Saitama;‡Institute for Virus Research, Kyoto University, Sakyo-ku, Kyoto; and

§Laboratory of Cell Regulation, School of Pharmaceutical Sciences, Nagasaki University, Nagasaki, Japan

Abstract: We examined the potential neurotrophic ef-fects of bone morphogenetic protein (BMP)-2 on thesurvival and differentiation of neurons cultured from therat developing striatum at embryonic day 16, a periodduring which the mRNAs for BMP-2 and its receptorsubunits (types IA, IB, and II) were detected. BMP-2exerted potent activity to promote the survival of striatalneurons and increased the number of surviving microtu-bule-associated protein-2-positive cells by 2.4-fold ascompared with the control cultures after 4 days in vitro.Although basic fibroblast growth factor (bFGF) alsoshowed relatively high activity to promote the survival ofstriatal neurons, transforming growth factor-b1, -b2, and-b3, glial cell line-derived neurotrophic factor, or brain-derived neurotrophic factor promoted their survivalweakly. Striatal neurons cultured in the presence ofBMP-2 or bFGF possessed extensive neurite outgrowths,the majority of which were GABA-immunoreactive. Inhi-bition of glial cell proliferation by 5-fluorodeoxyuridine didnot affect the capacity of BMP-2 to promote the survivalof striatal GABAergic neurons. In contrast, the ability ofbFGF to promote the survival of striatal neurons wasinhibited significantly by the treatment of cells with 5-flu-orodeoxyuridine. All these results suggest that BMP-2exerts potent neurotrophic effects on the striatalGABAergic neurons in a glial cell-independent manner.Key Words: Bone morphogenetic protein-2—Transform-ing growth factor-b superfamily—Neurotrophic factor—Striatal GABAergic neuron.J. Neurochem. 72, 2264–2271 (1999).

Bone morphogenetic proteins (BMPs) are a family ofproteins originally identified and characterized by theirability to induce cartilage and bone formation in ectopicextraskeletal sites in vivo (Wozney et al., 1988). Re-cently, several members of this protein family have beencloned, expressed in mammalian cells, and characterized.Comparative amino acid sequence analysis of BMPs hassuggested that these molecules can be divided into sev-eral groups of closely related proteins (Celeste et al.,

1990; Kingsley, 1994; Hogan, 1996). For example,BMP-2, BMP-4, and DPP (Drosophiladecapentaplegicprotein) form one subgroup, BMP-5, BMP-6 (Vgr-1),BMP-7 (OP-1), BMP-8 (OP-2), and 60A form anothersubgroup, and BMP-3 and growth and differentiationfactor-10 form a discrete subgroup of the family.

All the BMPs contain the characteristic seven highlyconserved cysteines in their carboxy-terminal portionsand thus belong to the transforming growth factor(TGF)-b superfamily, which includes TGF-bs, activins,inhibins, and glial cell line-derived neurotrophic factor(GDNF). It is becoming evident that all these TGF-bsuperfamily members are multifunctional proteins thatare implicated in the regulation of a wide range ofbiological phenomena, such as cell proliferation, celldifferentiation, immunosuppression, inflammation, tissuerepair, and embryogenesis (Massague, 1990; Kingsley,1994; Moses and Serra, 1996). Although BMPs playimportant roles in normal bone growth, bone remodeling,and fracture repair, the expression of BMPs and theirreceptors in many tissues other than bone (Iwasaki et al.,1995) suggests that they are also involved in the regula-tion of many biological processes unrelated to boneformation. Increasing evidence suggests a regulatory rolefor BMPs in early vertebrate embryogenesis, which in-

Received November 9, 1998; final revised manuscript received Jan-uary 14, 1999; accepted January 14, 1999.

Address correspondence and reprint requests to Dr. M. Kohno atLaboratory of Cell Regulation, School of Pharmaceutical Sciences,Nagasaki University, 1-14, Bunkyo-machi, Nagasaki 852-8131, Japan.

The present address of Dr. M. Katayama is Department of Neuro-physiology, Institute for Brain Research, University of Tokyo, Hongo,Tokyo 113-8655, Japan.

Abbreviations used:BDNF, brain-derived neurotrophic factor;bFGF, basic fibroblast growth factor; BMP, bone morphogenetic pro-tein; E, embryonic day; 5-FdUR, 5-fluorodeoxyuridine; GDNF, glialcell line-derived neurotrophic factor; MAP-2, microtubule-associatedprotein-2; NT, neurotrophin; PBS, phosphate-buffered saline; SN, sub-stantia nigra; TGF, transforming growth factor.

2264

Journal of NeurochemistryLippincott Williams & Wilkins, Inc., Philadelphia© 1999 International Society for Neurochemistry

cludes mesoderm induction, limb development, and he-matopoietic formation (Lyons et al., 1990; Jones et al.,1991; Storm et al., 1994; Johansson and Wiles, 1995;Hogan, 1996). Moreover, BMPs have gained attentionrecently in the study of CNS development; BMPs areexpressed in multiple CNS regions throughout develop-ment (Graham et al., 1994; Schluesener and Meyermann,1994; Tomizawa et al., 1995). In this respect, we havedemonstrated that BMP-2 acts as a neurotrophic factorand induces the neuronal differentiation of rat pheochro-mocytoma PC12 cells (Iwasaki et al., 1996). BMPs havealso been reported to inhibit initially the formation ofneuroectoderm during gastrulation, whereas within theneural tube they act as morphogens to promote the dif-ferentiation of dorsal cell types and intermediate celltypes (Hemmati-Brivanlou and Melton, 1997; Mehler etal., 1997) and to promote survival and subsequent dif-ferentiation of midbrain dopaminergic neurons and as-troglial cells (Gross et al., 1996; Jordan et al., 1997;Mabie et al., 1997).

Midbrain dopaminergic neurons are scattered through-out the ventral part of the tegmentum, which includes thesubstantia nigra (SN), the ventral tegmental area, andother adjacent nuclei, and form a continuum of dopami-nergic cells. Although these dopaminergic neurons haveascending, descending, and local connections, by far themost important pathway is the mesotelencephalic sys-tem, which is composed of the nigrostriatal, mesocorti-cal, and mesolimbic pathways (Fallon and Loughlin,1985). The caudate-putamen (striatum), one of the majortargets of the mesotelencephalic system, contains severaltypes of cells with different morphologies; the mostpopulous of which are the medium-sized spiny neurons(Heimer et al., 1985). Neurons of this type represent thecenterpiece of both intrinsic and extrinsic striatal cir-cuitry by receiving afferents from the SN dopaminergicneurons and projecting back to the SN, either directly orvia the globus pallidus. Both the strionigral and striopal-lidal efferent fibers are known to have GABA as theirneurotransmitter, although the former also conveys sub-stance P to the SN (Carpenter and Sutin, 1983). Anothertype of GABAergic neuron, characterized by a medium-sized aspiny morphology, contributes primarily to thestriatal intrinsic circuitry together with the cholinergicand somatostatinergic interneurons (Heimer et al., 1985).Thus, the striatal GABAergic neurons play a central rolein the reciprocal organization between the striatum andthe midbrain system.

In the present study, we sought to investigate whetheror not BMP-2 exerts neurotrophic effects on rat embry-onic striatal neurons. We show that BMP-2 markedlypromotes survival and differentiation of the striatalGABAergic neurons in culture. Although 5-fluorode-oxyuridine (5-FdUR), an antimitotic agent, apparentlysuppresses the proliferation of glial cells, it does notshow any inhibitory effect on the neurotrophic action ofBMP-2.

MATERIALS AND METHODS

Striatal cell cultureStriatal cell cultures were prepared as described previously

(Ishii et al., 1992). In brief, striata dissected from embryonicday 16 (E16) Wistar rats were minced in Ca21- and Mg21-freephosphate-buffered saline (PBS) and treated with 0.02% EDTAin PBS for 5 min at 37°C. Cells were dissociated by trituration,washed three times with serum-free Dulbecco’s modified Ea-gle’s medium, and suspended in serum-free defined medium(Dulbecco’s modified Eagle’s medium containing 20mg/mltransferrin, 25mM putrescine, 5 nM progesterone, 30 nMselenium, and 0.1 mg/ml bovine serum albumin). After beingpassed through a 200-mm nylon filter, cells were seeded at adensity of 53 104/cm2 in 24-well culture plates coated withpoly-D-lysine and incubated for 4 days. Cell cultures preparedas described above contained neuronal cells as;99% of thecell population, which were immunoreactive for microtubule-associated protein-2 (MAP-2);,1% of the cells were immu-noreactive for vimentin, a marker for glial cells (see Fig. 6B).

ImmunohistochemistryStriatal cultures were fixed in 4% paraformaldehyde for 30

min at room temperature and then washed three times withPBS. After treatment with 0.3% H2O2, cells were preincubatedin blocking buffer (PBS containing 0.3% Triton X-100 and 5%normal goat or horse serum) for 2 h at room temperature,followed by overnight incubation in primary antibody at 4°Cwith gentle agitation. The following antibodies were used:rabbit anti-GABA antibody (1:150; Chemicon), mouse anti-MAP-2 antibody (1:1,000; Amersham), and mouse anti-vimen-tin antibody (1:500; Sigma). After the cultures were washedthree times with PBS containing 0.3% Triton X-100 andblocked with 1% normal goat serum, they were incubatedfurther with secondary antibody (biotinylated goat anti-rabbitIgG or horse anti-mouse IgG) (1:200; Vector Laboratories) for90 min at room temperature. Immunoreactive cells were visu-alized after incubation with avidin-horseradish peroxidase anddevelopment with diaminobenzidine. No staining was observedin the control experiments in which primary antibodies wereomitted.

Measurement of the biological effects of BMP-2 onstriatal neurons

Isolated neurons were incubated for 4 days in the serum-freedefined medium with or without growth factors and then pro-cessed for GABA or MAP-2 immunoreactivity. After eachculture was photographed, the survival of neuronal cells wasevaluated by counting representative fields of view making upone quarter of the total culture dish area and then multiplyingthe result by 4. Neurite-outgrowth promotion activity wasquantified by counting cells bearing neurites longer than threetimes the length of the cell body (positive cells). A total of 100cells per well were chosen randomly, and positive cells werecounted; each data point corresponds to the counts obtainedfrom three independent wells.

RT-PCRTwo micrograms of total RNA isolated form the dissected

striatum with TRIzol (GibcoBRL) was digested with DNase Ito eliminate genomic DNA and then reverse-transcribed in avolume of 20ml with oligo-dT18 primer (70 ng) and Super-script II (GibcoBRL) reverse transcriptase according to themanufacturer’s instruction. The resulting cDNA was amplifiedby PCR with primers specific for the different BMPs or BMPreceptors. Oligonucleotide primers designed for murine BMP-2

J. Neurochem., Vol. 72, No. 6, 1999

2265NEUROTROPHIC EFFECT OF BMP-2 ON STRIATAL NEURONS

(sense: 59-TCAATTTAAGTTCTGTCCCTACTG-39, andantisense: 59-TCTTACCTGAGACCAGCTGTGTTCAT-39),murine BMP-4 (sense: 59-TAACCTCAGCAGCATCCCA-GAGA-39, and antisense: 59-CCGGAGTTGGGCCCAATT-TCCACTCC-39), BMP receptor subunits murine type IA(sense: 59-GAGGATCCTCTCATGTTCAAGGGCAGAA-39, and antisense: 59-GAGGATCCTTACATCAGATTAC-TG-39), murine type IB (sense: 59-GAGGATCCACAA-GAGAGCAAACAAAAAG-3 9, and antisense: 59-GAG-GATCCGCATTTACAACGTAGGATCT-39), and humantype II (sense: 59-GAGGATCCTGCCCTGGTGATTCTTG-GCTG-39, andantisense:59-GAGGATCCTCATGAGAGAT-TCTACTCTCACC-39) were examined (Koenig et al., 1994;Liu et al., 1995; Rosenzweig et al., 1995).

The PCR was carried out in a 100-ml volume containing 2.5ml of cDNA, 0.25mM primers, 200mM dNTPs, 13 AmpliTaqbuffer, and 1.5 mM MgCl2. After the addition of AmpliTaqGold (2.5 units) (Perkin-Elmer), mixtures were preincubated at95°C for 10 min and then the cDNAs were amplified with 35cycles of 94°C for 30 s, 45°C for 30 s, and 72°C for 45 s. ThePCR products were then separated by agarose gel electropho-resis.

MaterialsRecombinant human BMP-2, expressed in silkworm larvae,

was purified to homogeneity as described previously (Ishida etal., 1994). Recombinant human TGF-b1 and TGF-b2 werepurchased from Boehringer Mannheim GmbH (Mannheim,Germany); recombinant human TGF-b3 was from OncogeneScience, Inc. (Cambridge, MA, U.S.A.); recombinant humanGDNF, recombinant human brain-derived neurotrophic factor(BDNF), and basic fibroblast growth factor (bFGF) were fromToyobo Co. (Osaka, Japan). Recombinant human insulin, bo-vine transferrin, and poly-D-lysine were from Sigma (St. Louis,MO, U.S.A.). Other chemicals and reagents were of the purestgrade available.

RESULTS

Expression of BMP-2 and its receptor subunits inE16 rat striatum

Before studying the putative roles of BMP-2 in thedevelopment of striatum, we analyzed by RT-PCRwhether or not the mRNAs for BMP-2 and its receptorsubunits are expressed in the embryonic rat striatum. Asshown in Fig. 1, BMP-2 mRNA and also BMP-4 mRNAwere expressed in the developing rat striatum at E16.mRNAs for the BMP receptor subunit types IA, IB, andII were also detected in the E16 rat striatum, indicatingthe formation in the striatum of active receptor het-erodimers capable of mediating BMP signals. All thesemRNA species were detected in the E14 rat striatum(data not shown). Although the exact cell populationsexpressing BMP ligands and their receptors are obscureat present, these results suggested that BMP-2 could playroles in the development of the striatum.

BMP-2 promotes survival and differentiation ofstriatal neurons

We screened BMP-2 and several growth factors fortheir capacity to promote survival of striatal neurons;some of these factors have been reported previously tohave survival-promoting activity for the midbrain and/or

striatal neurons (Knu¨sel et al., 1990; Poulsen et al., 1994;Ventimiglia et al., 1995; Hou et al., 1996). Thus, cellsderived from E16 rat striatum were plated at low density(50,000 cells/cm2) in the serum-free defined medium andincubated for 4 days with or without growth factors. Theneuronal purity of the cultures were assessed by deter-mining the number of cells immunoreactive for MAP-2.Typically, our control cultures contained MAP-2-posi-tive cells whose number represented;99% of the cellpopulation after a 4-day incubation (Fig. 2). Only a smallpopulation of the cells was vimentin-positive and had aflattened, phase-dark appearance, which clearly distin-guished them from neuronal cells.

BMP-2 clearly increased the number of surviving stri-atal neurons (MAP-2-positive cells) by 2.4-fold com-pared with control cultures after 4 days in vitro: thenumber of MAP-2-positive cells represented 96.76 0.32% of the cell population, whereas that of vimen-tin-positive cells was 2.236 0.09% (n5 3 independentcultures) (Figs. 2 and 3). TGF-b1, TGF-b2, TGF-b3, andGDNF, members of the TGF-b superfamily, and BDNF,a member of a family of neurotrophins (NTs) structurallyrelated to nerve growth factor, showed a moderate pro-moting effect on the survival of MAP-2-positive cells(1.2–1.4-fold of the control). bFGF, which has beenshown to exert a neurotrophic effect on mesencephalicdopaminergic neurons, albeit indirectly through glialcells (Engele and Bohn, 1991), showed relatively highactivity to promote the survival of striatal neurons (two-fold of the control): the number of MAP-2-positive cellsrepresented 82.16 2.13% of the cell population, and thatof vimentin-positive cells represented 20.56 3.34% (n5 3). The survival-promoting activity of each growthfactor on the neurons was dose-dependent, and the max-imally effective concentrations are shown in Fig. 2 (seealso Fig. 4). BMP-2 and bFGF also showed marked

FIG. 1. Expression of BMP-2 and its receptors in the E16 ratstriatum. Total RNA obtained from the E16 rat striatum wasamplified by RT-PCR for 35 cycles and separated by agarose gelelectrophoresis as described in Materials and Methods. BMPR-IA, -IB, and -II, BMP receptor subunit types IA, IB, and II, re-spectively.


2266 A. HATTORI ET AL.

survival-promoting activity on neuronal cells obtainedfrom the E14 rat striatum (data not shown).

In addition to the promotion of striatal neuron sur-vival, BMP-2 induced morphological changes in theneurons (Fig. 2). Although control cultures containedround-shaped neurons with only short neurites, all theneuronal cells cultured in the presence of BMP-2 pos-sessed extensive neurite outgrowths, which started toextend within the initial 1–2 days after BMP-2 treatment.In striatal neurons, bFGF was found to induce moderateneurite-outgrowth formation, and TGF-b1, TGF-b2,TGF-b3, GDNF, and BDNF were found to induce weakneurite-outgrowth formation.

To examine further the positive relationship betweenthe survival-promoting activity and the neurite out-

growth-inducing activity of BMP-2, striatal neuronswere treated with varying concentrations of BMP-2 for 4days. As shown in Fig. 4, as little as 0.01 ng/ml BMP-2was found to be active in promoting both the survivaland the neurite-outgrowth induction in striatal neurons,and 1 ng/ml BMP-2 was the maximal effective dose forboth activities. All these results demonstrated thatBMP-2 had the capacity to promote both survival anddifferentiation of the striatal neurons in vitro.

The neurotrophic effect of BMP-2 on striatalneurons is independent of glial cell proliferation

It has been reported recently that BMPs possess theability to promote survival of midbrain dopaminergicneurons. However, this occurs through an indirect mech-

FIG. 2. Effect of various factors on the sur-vival and differentiation of striatal neurons.Striatal cell cultures obtained from E16 ratswere incubated for 4 days with no addition(A), 1 ng/ml BMP-2 (B), 1 ng/ml TGF-b1 (C),1 ng/ml TGF-b2 (D), 1 ng/ml TGF-b3 (E), 10ng/ml GDNF (F), 10 ng/ml bFGF (G), or 10ng/ml BDNF (H). Cells were then immuno-stained with a monoclonal antibody directedagainst MAP-2. Arrowheads indicate surviv-ing cells that were nonimmunoreactive forMAP-2. Scale bar 5 50 mm.



anism involving BMP stimulation of the maturation andproliferation of astroglia cells that produce certain neu-rotrophic factor(s) (Jordan et al., 1997). Thus, we nextexamined whether or not inhibition of glial cell prolifer-ation by 5-FdUR affects the ability of BMP-2 to promotesurvival of the striatal neurons. Treatment of striatal cellcultures with 5-FdUR inhibited the proliferation of glialcells in a dose-dependent manner, with an almost com-plete inhibition being observed at.30 mM. As higher

concentrations of 5-FdUR were toxic to neurons, striatalcell cultures obtained from E16 rats were treated with orwithout 30mM 5-FdUR and incubated for 4 days in thepresence of BMP-2 or bFGF. In the absence of 5-FdUR,both BMP-2 and bFGF caused about twofold increases inthe number of surviving neurons (Fig. 5). The addition of5-FdUR to the cultures inhibited partially the bFGF-induced survival of striatal neurons (Fig. 5). However,5-FdUR did not affect the capacity of BMP-2 to promotethe survival of striatal neurons. BMP-2 also stimulatedthe maturation and proliferation of glial cells (Fig. 6B),as had been reported previously (Gross et al., 1996;Mabie et al., 1997). However, the number of glial cellswas only limited in our striatal cell cultures, and 5-FdURcompletely inhibited the BMP-2-induced proliferation ofthem (Fig. 6B). These results clearly indicated that thetrophic effect of BMP-2 on striatal neurons was indepen-dent of the proliferation of glial cells. In contrast, thesurvival-promoting activity of bFGF on striatal neuronswas dependent, at least in part, on the proliferation ofglial cells, as is the case of midbrain dopaminergic neu-rons (Engele and Bohn, 1991).

The majority of neurons in the striatum are medium-sized with spiny projections and are known to use GABAas a neurotransmitter (Graybiel and Ragsdale, 1983). Inthis regard, the majority of neurons obtained from theE16 rat striatum were GABA-immunoreactive and weresusceptible to the neurotrophic effect of BMP-2 andbFGF (Fig. 6A). Furthermore, although the presence of5-FdUR did not affect the ability of BMP-2 to promotethe survival of GABA-positive neurons, it apparentlyinhibited the neurotrophic effect of bFGF on them.

DISCUSSION

The present study clearly demonstrates that BMP-2exerts potent neurotrophic effects on striatal GABAergic

FIG. 3. Effect of various factors on the survival of striatal neu-rons. Striatal cell cultures obtained from E16 rats were incubatedfor 4 days with no addition, 1 ng/ml BMP-2, 1 ng/ml TGF-bs, 10ng/ml GDNF, 10 ng/ml bFGF, or 10 ng/ml BDNF. Numbers ofsurviving neurons were counted after each culture was photo-graphed as described in Materials and Methods and are ex-pressed as percentages of control (10,462 6 510 cells/well).Results represent the averages 6 SE of three independent ex-periments. Statistical analysis was by ANOVA followed byFisher’s post hoc test: *p , 0.05 or **p , 0.001 versus control.

FIG. 4. Dose responses of BMP-2 to promote the survival anddifferentiation of striatal neurons. Striatal cell cultures obtainedfrom E16 rats were incubated with various concentrations ofBMP-2 for 4 days. Numbers of surviving neurons (gray columns)and those with neurite outgrowth (solid columns) were countedas described in Materials and Methods. The number of survivingneurons in the control cultures was 10,252 6 737 cells/well.Results represent the averages 6 SE of three independent ex-periments.

FIG. 5. Effect of 5-FdUR treatment on the capacity of BMP-2and bFGF to promote the survival of striatal neurons. Striatal cellcultures obtained from E16 rats were incubated for 4 days withno addition, 1 ng/ml BMP-2, or 10 ng/ml bFGF in the absence(gray columns) or presence (solid columns) of 30 mM 5-FdUR.The number of surviving neurons in the control culture was12,085 6 612 cells/well. Results represent the averages 6 SE ofthree independent experiments. Significant difference betweenbFGF-treated cultures in the absence and presence of 5-FdUR:*p , 0.001 (Student’s t test).



neurons, the predominant neuron type in the striatum(Graybiel and Ragsdale, 1983). BMP-2 increases thenumber of surviving neurons by 2.4-fold compared withcontrol cultures after 4 days in vitro (Figs. 2 and 3).BMP-2 also induces neurite outgrowth of striatal neurons(Figs. 2 and 6A). Many BMPs, including BMP-2, havebeen shown recently to exert neurotrophic effects onmidbrain dopaminergic neurons (Jordan et al., 1997). Inthis respect, although several other members of theTGF-b superfamily, such as TGF-b1, TGF-b2, TGF-b3,activin A, and GDNF, have also been reported to possessa neurotrophic activity on midbrain dopaminergic neu-rons (Poulsen et al., 1994; Krieglstein et al., 1995; Houet al., 1996), their effects on striatal neurons are not soapparent as compared with that of BMP-2. Thus, BMP-2

appears to be a unique TGF-b superfamily member fromthe viewpoint of its neurotrophic action: BMP-2 sharesneurotrophic activity on dopaminergic midbrain neuronswith many other members of the TGF-b superfamilyand, in addition, it specifically exerts notable neurotro-phic effects on striatal GABAergic neurons.

Although several BMPs have been shown to act aspotent neurotrophic factors on midbrain dopaminergicneurons, they also stimulate the proliferation and matu-ration of astroglial cells, and suppression of cell prolif-eration by 5-FdUR treatment abrogates their neurotro-phic effects (Jordan et al., 1997). Thus, Jordan et al.(1997) have concluded that the effects of BMPs onmidbrain dopaminergic neurons are indirect: BMPs ini-tially act on astroglial cells and/or their progenitors to

FIG. 6. A: Effect of 5-FdUR treat-ment on the capacity of BMP-2 andbFGF to promote the survivalof GABA-immunoreactive striatalneurons. Striatal cell cultures ob-tained from E16 rats were incu-bated for 4 days with no addition, 1ng/ml BMP-2, or 10 ng/ml bFGF inthe absence of presence of 30 mM5-FdUR. Cells were then immuno-stained with a polyclonal antibodydirected against GABA. B: Effect of5-FdUR treatment on the capacityof BMP-2 to promote the prolifera-tion of glial cells. Striatal cell cul-tures obtained from E16 rats wereincubated for 4 days with 1 ng/mlBMP-2 in the absence (a) or pres-ence (b) of 30 mM 5-FdUR. Cellswere then immunostained witha polyclonal antibody directedagainst vimentin. Scale bar 5 50mm.



stimulate the production of certain trophic factor(s),which then act on the neurons. It has also been suggestedthat neurotrophic effects of several mitogenic growthfactors, such as bFGF, TGF-a, epidermal growth factor,and insulin-like growth factor, on midbrain dopaminer-gic neurons are indirect, as their effects are always ac-companied by marked increases in astroglial cell num-bers and are abolished upon inhibition of cell prolifera-tion (Knusel et al., 1990; Alexi and Hefti, 1993). Incontrast, several TGF-b superfamily members, such asTGF-b1, TGF-b2, TGF-b3, activin A, and GDNF, havebeen reported to share neurotrophic activity on dopami-nergic neurons without any need for the indirect media-tion of astroglial cells (Krieglstein et al., 1995). Neuro-trophins, such as BDNF, NT-3 and NT-4/5, have alsobeen shown to promote survival and/or differentiation ofstriatal neurons (Widmer and Hefti, 1994; Ventimiglia etal., 1995). As astrocytes do not express active receptorsfor NTs, the effects of these NTs on striatal neurons aresuggested to be direct and not to be mediated by theproliferation of astrocytes.

Although BMP-2 induces a small increase in glial cellnumber, suppression of glial cell proliferation by5-FdUR (Fig. 6B) has little effect on the neurotrophicactivity of BMP-2 on striatal neurons. In contrast, theability of bFGF to promote survival of striatal neurons isinhibited significantly by the suppression of glial cellproliferation by 5-FdUR. bFGF also increases the uptakeof GABA into striatal neurons, which is inhibited by thetreatment of cells with 5-FdUR (data not shown). Allthese results indicate that BMP-2 promotion of the sur-vival and differentiation of striatal GABAergic neuronsis unlikely to occur indirectly via glial cell proliferation.Rather, it seems likely that BMP-2 exerts its neurotro-phic effects directly on the striatal neurons, as has beenreported previously for TGF-b1 TGF-b2, TGF-b3, ac-tivin A, and GDNF on dopaminergic neurons (Kriegl-stein et al., 1995). The reason why BMP-2 exerts itsneurotrophic effects directly on GABAergic neurons andindirectly on dopaminergic neurons could be, for exam-ple, because GABAergic but not dopaminergic neuronsexpress functional BMP receptors and/or signal-trans-ducing molecules that function downstream of the recep-tors. It could also be possible that although BMP-2 actson GABAergic neurons and induces the production oftrophic factor(s), which then act on the cells in an auto-crine fashion, dopaminergic neurons do not producethese factors by themselves in response to BMP-2 stim-ulation, but require them to be produced by nonneuronalcells (such as glial cells) to promote their survival anddifferentiation; these possibilities remain to be examinedin a future study.

Growing evidence has suggested that many of theTGF-b superfamily members, including BMP-2, playkey roles in the initial stages of neurogenesis and orga-nogenesis during development (Kingsley, 1994; Mosesand Serra, 1996; Hemmati-Brivanlou and Melton, 1997).In this context, it may be worth mentioning that whereasBMP-2 induces neurogenesis in neuronal crest stem

cells, TGF-b promotes smooth muscle differentiation:stem cells seem to determine their fates by using signalsdelivered by different members of the TGF-b superfam-ily (Shah et al., 1996). Our results showing the presenceof both BMP-2 and its receptors in E16 rat striatum (Fig.1) suggest that BMP-2 could indeed contribute to stria-tum development as a result of its ability to promote thesurvival and differentiation of GABAergic neurons andalso by its ability to promote the differentiation andmaturation of astroglial cells (Gross et al., 1996; Mabieet al., 1997).

BMP-2 exerts its neurotrophic effects both on mid-brain dopaminergic neurons (although through an indi-rect mechanism) (Jordan et al., 1997) and on striatalGABAergic neurons. These two populations of neuronsare interconnected reciprocally and form well-character-ized pathways, such as nigrostriatal, strionigral, andstriopallidonigral fibers (Fallon and Loughlin, 1985). Asthese pathways contribute to the regulation of movement,their dysfunction results in various types of involuntarymovement. Thus, involvement of BMP-2 in the regula-tion of the survival and differentiation of these two typesof neurons suggests that BMP-2 plays an important rolein the development of reciprocal organization of mid-brain and striatum. In addition, disorders of the BMP-2signaling system and/or unbalanced regulation of bothtypes of neurons by BMP-2 may result in aberrant for-mation of the mesostriatal interconnection or failure tomaintain it, either of which would cause some abnormal-ity in movement. Huntington’s disease is one such move-ment disorder, which is characterized clinically by cho-reic movement and pathologically by a progressive de-generation of striatal GABAergic neurons (Reiner et al.,1988). Based on the ability of BMP-2 to promote thesurvival of striatal GABAergic neurons in culture,BMP-2 may protect striatal neurons from degeneration inthe disease state. This challenging possibility shouldprovide a focus for future study.

Acknowledgment: We are grateful to Dr. Atsuyoshi Shi-mada (Institute for Developmental Research, Aichi HumanService Center) for helpful discussion and advice, and to Dr.Patrick Hughes for critical reading of the manuscript. Thiswork was supported in part by grants-in-aid from the Ministryof Education, Science, Sports and Culture of Japan (to M.K.)and a grant for “Multibioprobe Research Program” fromRIKEN (to M.T.).

REFERENCESAlexi T. and Hefti F. (1993) Trophic actions of transforming growth

factor a on mesencephalic dopaminergic neurons developing inculture.Neuroscience55, 903–918.

Carpenter M. B. and Sutin J. (1983)Human Neuroanatomy,Williams& Wilkins, Baltimore.

Celeste A. J., Iannazzi J. A., Taylor R. C., Hewick R. M., Rosen V.,Wang E. A., and Wozney J. M. (1990) Identification of transform-ing growth factorb family members present in bone-inductiveprotein purified from bovine bone.Proc. Natl. Acad. Sci. USA87,9843–9847.

Engele J. and Bohn M. C. (1991) The neurotrophic effects of fibroblastgrowth factors on dopaminergic neurons in vitro are mediated bymesencephalic glia.J. Neurosci.11, 3070–3078.



Fallon J. H. and Loughlin S. E. (1985) Substantia nigra, inThe RatNervous System, Vol. 1(Paxinos G., ed), pp. 353–374. AcademicPress, Sydney.

Graham A., Francis-West P., Brockell P., and Lumsden A. (1994) Thesignaling molecule BMP4 mediates apoptosis in the rhombence-phalic neural crest.Nature15, 684–686.

Graybiel A. M. and Ragsdale C. W. J. (1983) Biochemical anatomy ofthe striatum, inChemical Neuroanatomy(Emson P. C., ed), pp.427–504. Raven Press, New York.

Gross R. E., Mehler M. F., Mabie P. C., Zang Z., Santschi L., andKessler J. A. (1996) Bone morphogenetic proteins promote astro-glial lineage commitment by mammalian subventricular zone pro-genitor cells.Neuron17, 595–606.

Heimer L., Alheid G. F., and Zaborszky L. (1985) Basal ganglia, inTheRat Nervous System, Vol. 1(Paxinos G., ed), pp. 37–86. Aca-demic Press, Sydney.

Hemmati-Brivanlou A. and Melton D. (1997) Vertebrate embryoniccells will become nerve cells unless told otherwise.Cell 88,13–17.

Hogan B. L. M. (1996) Bone morphogenetic proteins: multifunctionalregulators of vertebrate development.Genes Dev.10,1580–1594.

Hou J.-G. G., Lin L.-F. H., and Mytilineou C. (1996) Glial cellline-derived neurotrophic factor exerts neurotrophic effects ondopaminergic neurons in vitro and promotes their survival andregrowth after damage by 1-methyl-4-phenylpyridinium.J. Neu-rochem.66, 74–82.

Ishida N., Tsujimoto M., Kanaya T., Shimamura A., Tsuruoka N.,Kodama S., Katayama T., Oikawa S., Matsui M., Nakanishi T.,Kobayashi J., and Nakazato H. (1994) Expression and character-ization of human bone morphogenetic protein-2 in silkworm lar-vae infected with recombinantBombyx morinuclear polyhedrosisvirus. J. Biochem. (Tokyo)115,279–285.

Ishii K., Inoue H., Hakura A., Araki W., and Masuda T. (1992)Immortalization of fetal mouse brain glial cells by human papil-lomavirus type 16 E7 genes.Cell Struct. Funct.17, 197–202.

Iwasaki S., Tsuruoka N., Hattori A., Sato M., Tsujimoto M., and KohnoM. (1995) Distribution and characterization of specific cellularbinding proteins for bone morphogenetic protein-2.J. Biol. Chem.270,5476–5482.

Iwasaki S., Hattori A., Sato M., Tsujimoto M., and Kohno M. (1996)Characterization of the bone morphogenetic protein-2 as a neuro-trophic factor: induction of neuronal differentiation of PC12 cellsin the absence of mitogen-activated protein kinase activation.J. Biol. Chem.271,17360–17365.

Johansson B. M. and Wiles M. V. (1995) Evidence for involvement ofactivin A and bone morphogenetic protein 4 in mammalian me-soderm and hematopoietic development.Mol. Cell. Biol.15,141–151.

Jones C. M., Lyons K. M., and Hogan B. L. M. (1991) Involvement ofbone morphogenetic protein-4 (BMP-4) and Vgr-1 in morphogen-esis and neurogenesis in the mouse.Development111,531–542.

Jordan J., Bottner M., Schluesener H. J., Unsicker K., and KrieglsteinK. (1997) Bone morphogenetic proteins: neurotrophic roles ofmidbrain dopaminergic neurons and implications of astroglialcells.Eur. J. Neurosci.9, 1699–1710.

Kingsley D. M. (1994) The TGF-b superfamily: new members, newreceptors, and new genetic tests of function in different organisms.Genes Dev.8, 133–146.

Knusel B., Michel P. P., Schwaber J., and Hefti F. (1990) Selective andnon-selective stimulation of central cholinergic and dopaminergicdevelopment in vitro by nerve growth factor, basic fibroblastgrowth factor, epidermal growth factor, insulin and the insulin-like growth factors I and II.J. Neurosci.10, 558–570.

Koenig B. B., Cook J. S., Wolsing D. H., Ting J., Tiesman J. P., CorreaP. E., Olson C. A., Pedquet A. L., Ventura F., Grant R. A., ChenG.-X., Warana J. L., Massague J., and Rosenbaum J. S. (1994)Characterization and cloning of a receptor for BMP-2 and BMP-4from NIH 3T3 cells.Mol. Cell. Biol. 14, 5961–5974.

Krieglstein K., Suter-Crazzolara C., Fisher W. H., and Unsicker K.(1995) TGF-b superfamily members promote survival of midbraindopaminergic neurons and protect them against MPP1 toxicity.EMBO J.14, 736–742.

Liu F., Ventura F., Doody J., and Massague J. (1995) Human type IIreceptor for bone morphogenetic proteins (BMPs): extension oftwo-kinase receptor model to the BMPs.Mol. Cell. Biol. 15,3479–3486.

Lyons K. M., Pelton R. W., and Hogan B. L. M. (1990) Organogenesisand pattern formation in the mouse; RNA distribution patternssuggest a role for bone morphogenetic protein-2A (BMP-2A).Development109,833–844.

Mabie P. C., Mehler M. F., Marmur R., Papavasiliou A., Song Q., andKessler J. A. (1997) Bone morphogenetic proteins induce astro-glial differentiation of oligodendroglial–astroglial progenitorcells.J. Neurosci.17, 4112–4120.

Massague J. (1990) The transforming growth factor-b family. Annu.Rev. Cell Biol.6, 597–641.

Mehler M. F., Mabie P. C., Zhang D., Kessler J. A. (1997) Bonemorphogenetic proteins in the nervous system.Trends Neurosci.20, 309–317.

Moses H. L. and Serra R. (1996) Regulation of differentiation byTGF-b. Curr. Opin. Genet. Dev.6, 581–586.

Poulsen K. T., Armanini M. P., Klein R. D., Hynes A., Phillips H. S.,and Rosenthal A. (1994) TGF-b2 and TGF-b3 are potent survivalfactors for midbrain dopaminergic neurons.Neuron 13, 1245–1252.

Prehn J. H., Backhauss C., and Krieglstein J. (1993) Transforminggrowth factor-b1 prevents glutamate neurotoxicity in rat neocor-tical culture and protects mouse neocortex from ischemic injury invivo. J. Cereb. Blood Flow Metab.13, 521–525.

Reiner A., Albin R. L., Anderson K. D., D’Amato C. J., Penney J. B.,and Young A. (1988) Differential loss of striatal projection neu-rons in Huntington disease.Proc. Natl. Acad. Sci. USA85,5733–5737.

Rosenzweig B. L., Imamura T., Okadome T., Cox G. N., Yamashita H.,ten Dijke P., Heldin C.-H., and Miyazono K. (1995) Cloning andcharacterization of a human type II receptor for bone morpho-genetic proteins.Proc. Natl. Acad. Sci. USA92, 7632–7636.

Schluesener H. J. and Meyermann R. (1994) Expression of BMP-6, aTGF-b related morphogenetic cytokine, in rat radial glial cells.Glia 12, 161–164.

Shah N. M., Groves A. K., and Anderson D. J. (1996) Alternativeneural crest cell fates are instructively promoted by TGF-b super-family members.Cell 85, 331–343.

Storm E. E., Huynh T. V., Copeland N. G., Jenkins N. A., Kingsley D.,and Lee S.-J. (1994) Limb alterations inbrachypodismmice dueto mutations in a new member of the TGFb-superfamily.Nature368,639–642.

Tomizawa K., Matsui H., Kondo E., Miyamoto K., Tokuda M., ItanoT., Nagahata S., Akagi T., and Hatase O. (1995) Developmentalalteration and neuron-specific expression of bone morphogeneticprotein-6 (BMP-6) mRNA in rodent brain.Brain Res. Mol. BrainRes.28, 122–128.

Ventimiglia R., Mather P. E., Jones B. E., and Lindsay R. M. (1995)The neurotrophins BDNF, NT-3 and NT-4/5 promote survival andmorphological and biochemical differentiation of striatal neuronsin vitro. Eur. J. Neurosci.7, 213–222.

Widmer H. R. and Hefti F. (1994) Neurotrophin-4/5 promotes survivaland differentiation of rat striatal neurons developing in culture.Eur. J. Neurosci.6, 1669–1679.

Wozney J. M., Rosen V., Celeste A. J., Mitscock L. M., Whitters M. J.,Kriz R. W., Hewick R. M., and Wang E. A. (1988) Novelregulators of bone formation: molecular cloning and activities.Science242,1528–1534.



bone morphogenetic protein-2 promotes survival and differentiation of striatal gabaergic neurons in...

Documents