Co-Grafted Embryonic StriatumIncreases the Survival of Grafted

Embryonic Dopamine Neurons

CARYL E. SORTWELL,1* TIMOTHY J. COLLIER,2 AND JOHN R. SLADEK, JR.1

1Department of Neuroscience, Chicago Medical School, North Chicago, Illinois 600642Department of Neurological Sciences, Rush Presbyterian St. Luke’s Medical School,

Chicago, Illinois 60612

ABSTRACTTo enhance the current therapeutic benefit of dopamine (DA) neuron grafts in Parkinson’s

disease, strategies must be developed that increase both DA neuron survival and fiberoutgrowth into the denervated striatum. Previous work in our laboratory has demonstratedthat dopaminergic neurons grow to greater size when co-grafted with striatal cell suspensionsand display extensive tyrosine hydroxylase-positive (TH1) projections, but no conclusioncould be reached concerning enhancement of survival of grafted DA neurons. The aim of thepresent study was to characterize further the potential trophic effects of striatal co-grafts ongrafted mesencephalic DA neuron survival. Unilaterally lesioned male Fischer 344 rats weregrafted with either a suspension of mesencephalic cells or with both mesencephalic andstriatal cell suspensions. Co-grafts were either mixed together or placed separately into thestriatum. Lesioned rats receiving no graft served as controls. Rotational behavior wasassessed following amphetamine challenge at 2 weeks prior to grafting and at 4 and 8 weeksfollowing grafting. Only rats receiving co-grafts of nigral and striatal suspensions separatedby a distance of 1 mm showed significant behavioral recovery from baseline rotationalasymmetry. Both mixed and separate striatal co-grafts were associated with a doubling of DAneuron survival compared with solo mesencephalic grafts. In the mixed co-graft experiment,DA neurite branching appeared enhanced and TH-rich patches were observed, whereas withco-grafts that were separated, TH1 innervation of the intervening host striatum wasincreased significantly. These results provide the first evidence suggesting that nigral-striatalco-grafts, particularly those placed separately and in proximity to each other, increase bothDA neuron survival and neurite extension from the mesencephalic component of the grafts.J. Comp. Neurol. 399:530–540, 1998. r 1998 Wiley-Liss, Inc.

Indexing terms: transplant; substantia nigra; Parkinson’s disease; trophic striatum; dopamine

The abnormal motor function that is symptomatic ofParkinson’s disease (PD) is not usually apparent untildopamine (DA) levels in the striatum have dropped to lessthan 20% of normal (Hornykiewicz, 1988). Attempts toincrease DA levels through the transplantation of pharma-cologically relevant cells has provided a novel therapeuticstrategy for PD. Implants of substantia nigra DA cells tothe caudate nucleus in monkeys significantly alleviatespecific behavioral deficits following 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) treatment (Taylor etal., 1990, 1991). However, human clinical studies withfetal tissue transplants have not resulted in similar suc-cess. With few exceptions, patients have experienced ei-ther no improvement or modest improvement (Ahlskog,1993; Olanow et al., 1996). In the most recent case study, inwhich all grafted patients experienced significant clinical

benefit (Freeman et al., 1995), tissue from six to eightdonors was transplanted to each recipient. For the trans-plant approach to become a significant therapeutic option,strategies must be developed to enhance the survival ofgrafted DA neurons.

The striatum provides a target for DA-containing inner-vation from the substantia nigra during development ofthe nigrostriatal pathway. Multiple influences, including

Grant sponsor: NIH; Grant numbers: F32 NS09646–01, RSA K05-MH00643, and AG10851.

*Correspondence to: Caryl E. Sortwell, Ph.D., Department of Neurologi-cal Sciences, Rush-Presbyterian-St. Luke’s Medical Center, Tech 2000,Suite 200, 2242 W. Harrison Street, Chicago, IL 60612.E-mail: csortwel@rush.rpslmc.edu

Received 14 April 1997; Revised 21 May 1998; Accepted 25 May 1998

THE JOURNAL OF COMPARATIVE NEUROLOGY 399:530–540 (1998)

r 1998 WILEY-LISS, INC.

extracellular matrix components and trophic factors, con-tribute to the establishment of this circuitry. There isabundant evidence in in vitro systems that striatal targettissue exerts both tropic and trophic effects on developingmesencephalic DAneurons (Prochiantz et al., 1979; Daguetet al., 1980; Hemmendinger et al., 1981; Hoffman et al.,1983; Tomozawa and Appel, 1986; Dal Toso et al., 1988;Niijima et al., 1990; Dong et al., 1993; Takeshima et al.;1994, Carvey et al., 1996). Attempts have been made toextend these effects observed in culture to the mammalianbrain. DA neurons survived and innervated their embry-onic striatal targets when co-grafted to the anterior cham-ber of the eye of adult rats (Olson et al., 1979). Greater hostreinnervation density was produced by intrastriatal graftsof mixed embryonic mesencephalic and striatal cell suspen-sions than by mesencephalic grafts alone (Brundin et al.,1986). Moreover, embryonic striatal grafts were reinner-vated preferentially over host striatum (DeBeaurepaireand Freed, 1987). Significantly larger DA cell bodies andextensive TH-positive (TH1) projections were seen whenboth mixed and separate striatal-nigral co-grafts werestudied, respectively (Yurek et al., 1990). Embryonic stria-tal tissue is effective in promoting TH-positive fiber out-growth from mesencephalic grafts placed homotopically inthe substantia nigra (Dunnett et al., 1989). A strongcorrelation exists between co-graft-enhanced TH1 fiberoutgrowth and more rapid or enhanced functional recoveryfrom denervation (Brundin et al., 1986; Dunnett et al.,1989; Yurek et al., 1990). More recently, mixed nigral-striatal grafts exhibited a more robust and long-lastingturning response in a bilaterally lesioned model (Costan-tini et al., 1994). In nonhuman primates, dopaminergicneurites displayed preferential growth from a separatenigral graft to the accompanying striatal graft (Sladek etal., 1993).

These studies illustrate the predictive validity of in vitroexperiments. Striatal cells serve as a chemotactic at-tractant and enhance neurite outgrowth from mesence-phalic DA neurons both in vitro and in vivo. However,enhanced survival of mesencephalic neurons has beenreported only in culture. The present experiment providesthe first direct evidence that co-grafted striatal cellsinfluence the directionality and density of DA neuriteextension and enhance mesencephalic DA neuron sur-vival. To investigate whether target-derived regulation ofDA neuron survival and neurite extension is contactdependent or is mediated by diffusible substances, both amixed and separate co-graft condition were included. Thisresearch has been presented previously in abstract form(Sortwell et al., 1996, 1997).

MATERIALS AND METHODS

Lesions and behavioral assessment

Thirty-five male Fischer 344 rats (200–225 g) wereused in this study. Protocols were approved by the Institu-tional Animal Care and Use Committee of the ChicagoMedical School. Stereotaxic injections of 6-hydroxydopa-mine (6-OHDA) were made unilaterally into the nigrostria-tal pathway of anesthetized rats (30 mg/kg, pentobarbital).Each rat received a total of 18 µg of 6-OHDA (in 0.9%NaCl/0.2% ascorbic acid) in two injections over a period of3 minutes each, one in the vicinity of the medial forebrainbundle (AP 24.3, ML 11.2, DV 27.5) and the other in thepars compacta of the substantia nigra (AP 24.8, ML 11.5,

DV 27.5). Baseline measures of amphetamine-inducedrotational behavior were obtained at 2 weeks following the6-OHDA lesion. Lesioned animals were given an injectionof 5.0 mg/kg (i.p.) amphetamine and placed in cylindricalbowls in which rotational behavior was quantified by aMacIntosh rotometer program starting 15 minutes afterinjection, for a period of 70 minutes. The rotation quantita-tion computer program and cylinders were designed byDennis Levin and generously provided by Dr. Barry Hoffer.Animals meeting the criterion of six ipsilateral turns perminute or greater were grafted 2 weeks later. Rats weretested for rotational asymmetry following amphetaminechallenge at 4 and 8 weeks after grafting. Completeness ofthe lesion was verified histologically at the conclusion ofthe experiment.

Dissection and transplantation procedures

Timed pregnant female Fischer 344 rats were injectedwith 5-bromo-28-deoxyuridine (BrdU) at 24 and 12 hours(50 mg in 50% EtOH/kg, i.p.) prior to tissue dissection tolabel dividing cells and allow for graft identification.Mesencephalic and striatal brain regions were dissectedby using sterile techniques from embryonic day (E15)fetuses and separately pooled in oxygenated, cold, sterilecalcium-magnesium-free buffer, pH 7.1 (CMF). Care wastaken to dissect out only the lateral ganglionic eminence(LGE), the primordial anlagen that develops into thestriatum (Pakzaban et al., 1993; Deacon et al., 1994). Cellsuspensions of embryonic mesencephalic or striatal tissuewere then prepared through a series of CMF rinses,incubated in 0.125% trypsin (Gibco, Gaithersburg, MD),rinsed in CMF again, and triturated in 0.004% DNase todisperse the cells into solution. This cell suspension methodgenerates a primary culture containing approximately3–5% tyrosine hydroxylase-positive (TH1) neurons (unpub-lished observations). Trypan blue was added to a sample ofcell suspension and viewed in a hemocytometer to assesscell viability and to determine cell counts. Cell suspensionsof greater than 95% viability were diluted for transplanta-tion to a concentration of 60,000 cells/µl in CMF. Cells wereloaded into a 10-µl syringe with a 25-gauge needle andattached to a stereotaxic needle holder for implantation.

Following behavioral screening, lesioned rats weregrafted with either (A) mesencephalic cell suspensionsalone (i.e., solo mesencephalic graft), (B) mixed striataland mesencephalic cell suspensions (i.e., mixed co-graft),or (C) separate striatal and mesencephalic cell suspen-sions (i.e., separate co-graft). Mesencephalic cell suspen-sions alone (group A) and mixed co-grafts (group B) weretransplanted into the dorsomedial region of the striatum(AP 10.7, ML 12.0, DV 26.5). In the separate co-graftgroup (group C), mesencephalic cells were transplanted tothe above coordinates with the striatal suspension trans-planted at a 1-mm distance lateral in the same coronalplane (AP 10.7, ML 13.0, DV 26.5). In all transplantgroups, equal numbers of the same mesencephalic suspen-sion were grafted to ensure initial grafting of an equiva-lent number of DA neurons. Approximately 180,000 cellsin a 3-µl volume were injected into each implantation site.In the case of the mixed co-graft, approximately 360,000cells (180,000 mesencephalon/180,000 striatum) in a 6-µlvolume were injected. Transplant surgeries were con-ducted in alternating order by group to prevent differencesin DA neuron viability caused by lengthy postdissection

STRIATAL CO-GRAFTS AND DOPAMINE NEURON SURVIVAL 531

intervals (Brundin et al., 1985). A control group of lesionedrats with no grafts was also included.

Immunohistochemical analysis

Histological evaluation of surviving, grafted DA neuronsand patterns of neurite outgrowth was performed immedi-ately following the completion of behavioral assessment(9 weeks). Rats were deeply anesthetized (50 mg/kg,pentobarbital, i.p.) and perfused intracardially with 4%paraformaldehyde in 0.1 M phosphate buffer, pH 7.4. Eachbrain was removed, postfixed for 24 hours, and immersedin 30% sucrose in phosphate buffer. Fixed brains weresliced on a freezing microtome (40 µm) and free-floatingbrain sections were immunostained by using antiseradirected against TH (Eugene Tech, Ridgefield Park, NJ;rabbit anti-bovine TH 1:10,000, overnight at 25°C). Sec-tions for immunostaining of bromodeoxyuridine (BrdU)-labeled nuclei were exposed to 0.1 N HCl for 30 minutesprior to primary incubation (Boehringer Manheim, India-napolis, IN; mouse anti-BrdU 1:500, overnight at 25°C).Triton X-100 was added to the phosphate buffer (0.3 % forTH, 0.8% for BrdU) during incubations and rinses topermeabilize cell and nuclear membranes. The Vector ABCdetection kit employing horseradish peroxidase stainingwas used following primary incubation. Manual countswere made at 2003 magnification in every third 40-µm-thick section of TH1 neurons identified within grafts ofeach graft recipient. The sum of these counts was thenadjusted according to the method of Abercrombie (1946).These cell counts were collected to provide informationconcerning relative survival of grafted TH1 neurons amongexperimental groups. Rats with grafts improperly placedin the ventral region of the striatum were only included inthe morphological analysis and not included in the behav-ioral analysis since variations in graft placement havebeen demonstrated to influence rotational recovery (Dun-net et al., 1981, 1983; Bjorklund et al., 1980).

Dopaminergic innervation density

Quantification of the density of TH1 striatal reinnerva-tion from grafted mesencephalon was performed by usingMocha computer imaging software (Jandel Scientific, SanRafael, CA). Fields of 200 3 300 µm (at 1003 magnifica-tion) immediately lateral to the graft site in all three graftconditions were measured in the three sections per animaldetermined to possess the highest number of surviving DAneurons. These values were compared with striatal densi-ties 2 mm lateral to graft sites, as well within unlesionedstriatum, lesioned nongrafted striatum, and TH-richpatches of mixed striatal-nigral co-grafts. Optical densitywithin the corpus callosum was measured to provide acontrol for nonspecific background staining. All measure-ments were taken by using the same light intensity andthreshold settings to exclude areas of zero density, i.e.,blood vessels.

Statistical analysis

Statistical analysis was performed by using the Super-ANOVA statistics package. ANOVA for repeated measureswas used to analyze the data obtained from rotationalassessment and ANOVA was used to analyze the data frommanual counts of DA neurons and TH1 innervationdensities. The Fisher’s Protected LSD post hoc test wasused to determine significant differences.

RESULTS

Rotational asymmetry

Twenty-nine of the 35 6-OHDA-lesioned rats exhibitedaverage ipsilateral turning rates of 6 turns per minute orgreater after amphetamine challenge. These rats subse-quently were assigned to either a transplant or controlgroup. There were no significant differences in baselinerotational scores between the different transplant groups(F [3, 20] 5 0.199; P $ 0.05). At 4 and 8 weeks followinggrafting only the separately co-grafted rats (n 5 6) dis-played average ipsilateral rotation rates that were signifi-cantly recovered compared with baseline (P # 0.05). Incontrast, solo grafted rats (n 5 8) did not display asignificant reduction in amphetamine-induced rotations at8 weeks following transplantation (P $ 0.05). Similarly,mixed striatal co-grafted rats (n 5 6) did not displaysignificant reductions in rotational asymmetry (P $ 0.05).Lesioned-only rats (n 5 5) showed a trend toward in-creased ipsilateral rotation rates at 8 weeks; however, thisincrease did not achieve significance (P $ 0.05). Thesebehavioral results are illustrated in Figure 1.

5-Bromo-28-deoxyuridine labeling

BrdU labeling of the mesencephalic and striatal suspen-sions allowed for clearer visualization of cells in thegrafted regions and proved to be useful in distinguishinggrafted striatal tissue from host striatal tissue. Within allgraft types, BrdU-labeled nuclei of different staining inten-sities were apparent (Fig. 2).

Fig. 1. Effect of striatal co-grafts on rotational asymmetry follow-ing amphetamine challenge. Groups of animals included: 6-OHDA-lesioned rats that received no graft (Lesion Only, n 5 5), mesence-phalic solo grafts (Solo Graft, n 5 8), mixed striatal-nigral co-grafts(Mixed Co-graft, n 5 6), and separate striatal co-grafts (Sep Co-graft,n 5 6). The average number of ipsilateral rotations per minutewere recorded over a 70-minute interval. Values represent the mean 6SEM for each group at 2 weeks following 6-OHDA lesion (Baseline)and 4 weeks and 8 weeks following transplantation. Significantdifferences from baseline rotational scores are denoted by asterisks(P # 0.05).

532 C.E. SORTWELL ET AL.

Mesencephalic graft morphology

Unilateral injection of 6-OHDA produced a near totalabsence of TH immunoreactivity within the lesioned hoststriatum. All grafted rats (n 5 24) contained visible graftswith numerous TH1 neurons. In solo mesencephalicgrafted rats (n 5 9), coronal sections containing TH1neurons spanned at least 240 µm in the rostral-caudal axisand were visualized typically as thin tracts of DA neuronssurrounded by a distinct halo of TH1 innervation in thehost striatum (Figs. 3A,D, 4A). Individual TH1 neuritescould be traced for a limited distance into the surroundinghost parenchyma (Fig. 4A). At higher magnifications, TH1neurons within the solo mesencephalic grafts appearedunipolar or bipolar, possessing a few apical dendritesand/or axonal branches (Fig. 5A).

Mixed striatal co-grafts (n 5 8) were typically dropletshaped and occupied a larger portion of the lesionedstriatum than the solo grafts. These grafts were found tospan at least 320 µm in the rostral-caudal axis. TH1neurons were found to be scattered throughout the graft inclusters, rather than the usual location around the graftperiphery (Figs. 3B,E, 4B). While a halo of TH1 innerva-tion was apparent in the mixed co-grafts, the majority ofTH1 fibers were visible within the co-graft rather than inthe host striatum (Fig. 4B). The most remarkable compo-nent of the mixed striatal co-grafts was the presence of

dense TH-immunoreactive patches that often appeared toform robust pericellular arrays (Fig. 4B). At higher magni-fications, DA neurons within the co-grafts appeared topossess numerous neuritic branches (Fig. 5B).

Separate striatal co-grafts (n 5 7) appeared longer in thedorsal-ventral axis than their solo counterparts. The mes-encephalic portion of the separate co-grafts spanned atleast 360 µm in the rostral-caudal axis. The striatalportion of the co-graft appeared as an unstained areaapproximately 800–1,000 µms lateral to the mesence-phalic portion and was only faintly distinguishable fromthe host striatum. DA neurons within the mesencephalicportion appeared to be present in clusters throughout thegraft (Figs. 3C,F, 4C). In a few co-grafts TH-immunoreac-tive patches were present in the striatal portion of theco-graft (Fig. 3C,F). A dense halo of TH1 innervation wasobserved interposed between the mesencephalic and stria-tal portion of the separate co-grafts (Fig. 3C,F). Thisinnervation into the surrounding and intervening hoststriatum appeared to be denser and also to innervate alarger host area than that seen in either the solo or mixedstriatal co-grafted rats. At higher magnifications, TH1neurites could be observed extending toward the regions ofthe striatal graft (Fig. 4C). In general, grafted DA neuronsin the separate co-grafts were similar in individual mor-phology to the DA neurons within the solo and mixedstriatal co-grafts (Fig. 5C).

Fig. 2. 5-Bromo-28-deoxyuridine (BrdU) labeling of grafted embry-onic mesencephalic and striatal cells. A: Solo mesencephalic graft (ng).B: Mixed striatal-nigral co-graft (mg). C: Separate nigral (ng) andstriatal (sg) co-graft. D: At higher magnification, individual BrdU-

labeled nuclei of different staining intensities are apparent (arrows) inthe striatal portion of the co-graft. H, host striatum. Scale bars 5 200µm in C (applies to A–C), 50 µm in D.

STRIATAL CO-GRAFTS AND DOPAMINE NEURON SURVIVAL 533

Survival of dopamine neurons

Mesencephalic grafts transplanted alone (n 5 9) dis-played an average of 481 surviving DA neurons. This valuerepresents an approximate survival rate of 5–9% of the DA

neurons initially transplanted. Separate striatal co-grafts(n 5 7) contained an average of 953 TH1 neurons,representing an approximate survival rate of 11–18%.Similarly, mixed striatal co-grafts (n 5 8) contained anaverage of 954 surviving DA neurons, a survival rate of

Fig. 3. Low-power views of solo mesencephalic grafts (A,D), mixedstriatal-nigral co-grafts (B,E) and separate striatal-nigral co-grafts(C,F) in the denervated rat striatum labeled with antisera againsttyrosine hydroxylase (TH). A,D: Coronal sections are shown throughsolo mesencephalic grafts containing a total of 640 and 222 dopamine(DA) neurons, respectively. B,E: Examples of mixed nigral-striatalco-grafts are depicted that possessed totals of 2,514 and 1,521 DAneurons. Dense TH-immunoreactive patches are apparent within the

mixed co-graft (asterisks). C,F: Two separate nigral-striatal co-graftsare shown. These two grafts contain totals of 1,143 and 1,050 DAneurons, respectively, in the nigral portions of the separate co-graft.Dense TH-immunoreactive patches are visible within areas in thestriatal portion of the separate co-graft (asterisks, sg) and robustTH-positive innervation of the intervening host striatum is alsoapparent (arrows). Scale bar 5 300 µm.

534 C.E. SORTWELL ET AL.

11–18%, although the number of surviving DA neuronswas more variable than in the separately co-grafted condi-tion. The increased DA neuron survival in the separateand mixed striatal co-graft groups was significant whencompared with survival rates of the solo mesencephalicgrafts (P # 0.05). Figure 6 illustrates the striatal co-grafteffects on DA neuron viability.

Extent of TH1 reinnervationof the host striatum

Dopaminergic reinnervation of the host striatum imme-diately adjacent to all graft types was significantly denser

than lesioned, nongrafted striatum. As suggested by quali-tative examination, the intervening striatum betweenseparately co-grafted striatum and mesencephalon con-tained a significantly denser innervation of TH1 fibersthan the striatum immediately adjacent to both the soloand the mixed co-graft groups (P # 0.01). This increaseddensity between separate striatal-nigral co-grafts re-flected a 50% restoration of dopaminergic innervationcompared with unlesioned striatum. When TH fiber inner-vation in fields further lateral (2 mm) to the graft sites wasexamined, no significant differences were noted whencompared with control measurements in lesioned, non-grafted striatum. Lastly, measurements taken within TH1fiber-rich patches of mixed co-grafts revealed a markedincrease in fiber density, reaching a level 40% greater thanunlesioned striatum. Figure 7 illustrates the differences instriatal dopaminergic reinnervation from the three grafttypes.

DISCUSSION

Previous evidence suggests that striatal co-grafts maybe used to stimulate neurite extension of grafted mesence-phalic DA neurons and enhance recovery of lesion-inducedrotational asymmetry. We now have demonstrated thatco-grafted striatal cells also exhibit a trophic effect ongrafted DA neurons reflected by increased DA neuronsurvival. Separate striatal co-grafts and mixed striatalco-grafts displayed a significant twofold increase in DAneuron survival when compared with solo mesencephalicgrafts. The present findings also confirm earlier reports inwhich striatal co-grafts placed at a short distance frommesencephalic grafts appear to attract dopaminergic neu-rites (Yurek et al., 1990), significantly enhancing innerva-tion of the intervening lesioned host striatum in theprocess. These morphological effects of striatal co-graftsappear to influence behavioral parameters directly. Sepa-rate co-grafts produced a significant reduction in rota-tional asymmetry at both 4 and 8 weeks after transplanta-tion. Neither solo grafts nor mixed striatal co-graftsdisplayed significant behavioral improvement at thesetime intervals.

Cell survival of DA neurons is enhanced when mesence-phalic neurons are co-cultured with striatal cells (Hoffmanet al., 1983; Dong et al., 1993), intact or denervated striatalextracts (Dal Toso et al., 1988; Niijima et al., 1990; Carveyet al., 1996), or striatal oligodendrocyte-type-2 astrocyte(O-2A) progenitor cells (Takeshima et al., 1994). Althoughthis effect has been observed in vitro, it had never beendocumented to occur in the co-graft situation prior to ourinitial reports (Sortwell et al., 1996, 1997). One reason forthis discrepancy could be the inclusion of nontarget tissuein the striatal portion of the co-graft in previous studies. Inthose co-graft studies (Olson et al., 1979; Brundin et al.,1986; DeBeaurepaire and Freed, 1987; Dunnett et al.,1989; Yurek et al., 1990; Sladek et al., 1993), a distinctionmight not have been made between the dissection of thelateral ganglionic eminence (LGE) to the exclusion of themedial ganglionic eminence (MGE). The MGE developsinto the globus pallidus (Pakzaban et al., 1993; Deacon etal., 1994), an inappropriate target for developing nigralneurons. Inclusion of the MGE in the striatal portion of theco-graft may have diluted the effects of LGE-derivedtrophic factors, particularly in cell suspension grafts wherethe number of co-grafted cells remains constant. Ourpresent results illustrating the trophic effect of striatal

Fig. 4. Effects of co-grafted striatal target tissue on the directional-ity and length of grafted mesencephalic dopamine (DA) neurites. A: Inthis example of a solo mesencephalic graft, tyrosine hydroxylase(TH)-immunoreactive neurites (arrows) can be seen extending only alimited distance into the host striatum. B: In mixed nigral-striatalco-grafts, whereas a halo of TH immunoreactivity is apparent alongthe border of the co-graft (open arrows), much denser TH innervation(asterisks) remains within the boundary of the graft (host-graft borderdemarcated by black on white arrows). C: Long individual DA neurites(black arrowheads) are readily visible extending from the nigralportion to the striatal portion of this separate co-graft. sg, striatalgraft. Scale bars 5 100 µm in A (applies to A and B), 50 µm in C.

STRIATAL CO-GRAFTS AND DOPAMINE NEURON SURVIVAL 535

co-grafts have been confirmed by a very recent manuscriptpublished during this manuscript’s review process. Mixednigral-LGE co-grafts were also demonstrated to possess anincreased survival of DA neurons (Costantini and Snyder-Keller, 1997).

Our behavioral and viability results appear to disagreewith the first of two studies conducted by Costantini et al.(1994). In this study, mixed striatal co-grafts did producean augmented turning response in a bilaterally lesionedmodel without significant effects on DA neuron viability. Infact, a trend toward decreased survival of DA neurons inthe mixed striatal-nigral grafts was reported. Inconsisten-cies between their results and ours may be due to method-ological differences: different embryonic age of the trans-planted tissue and a different behavioral model. Forexample, the bilateral 6-OHDA model employed in theirstudy allows for a larger range with which to measure theefficacy of the transplants. It is possible that the behav-ioral effects of mixed striatal co-grafts are too small to bedetected in the unilateral 6-OHDA model. The most strik-ing contrast between our experimental design and theirearlier design is the extreme difference in the number ofcells implanted. In their 1994 study a total of 500,000–1,000,000 cells were implanted compared with our 180,000–360,000 cells. Perhaps a larger graft volume is a subopti-mal grafting condition that offsets any trophic effects of thestriatal co-graft through increases in reactive gliosis andby limiting the access of grafted neurons to nutrients fromhost blood vessels. Nikkhah et al. (1994) found that themicrotransplantation of an equal amount of mesence-phalic cells in smaller multiple deposits enhances graftedDA neuron viability almost 300% and an experiment inwhich the vascularization of grafts was enhanced alsoreported increases in graft volume (Finger and Dunnett,1989). The most compelling evidence that the larger graftvolume used in the Costantini et al. (1994) study may haveproved detrimental to DA neuron survival is made avail-able by results of their second, more recent study. When

Fig. 5. High-power photographs depicting the individual cellularmorphology of dopamine (DA) neurons within solo mesencephalicgrafts (A), mixed striatal-nigral co-grafts (B), and separate striatal-nigral co-grafts (C). A: Tyrosine hydroxylase (TH)-positive neurons(arrows) in this solo mesencephalic graft possess few apical dendritesand/axonal branches. B: DA neurons within this mixed nigral-striatal

co-graft (arrows) are often obscured by dense neuritic branching andfiber ingrowth into TH-immunoreactive patches (asterisk). C: Neuriticbranching of DA neurons (arrows) within the nigral portion of thisseparate nigral-striatal co-graft appears similar to the branching ofDA neurons in the solo mesencephalic grafts depicted in A. Scale bar 5100 µm.

Fig. 6. Striatal co-grafts increase the survival of grafted mesence-phalic dopamine (DA) neurons. Groups of animals included: mesence-phalic solo grafts (Solo Graft, n 5 9), mixed striatal-nigral co-grafts(Mixed Co-graft, n 5 8), and separate striatal co-grafts (Sep Co-graft,n 5 7). Tyrosine hydroxylase (TH)-positive neurons were counted inevery third 40-µm section and the sum of these counts correctedaccording to the method of Abercrombie (1946). Values represent themean 6 SEM for each group at 9 weeks following transplantation.Significant differences from the solo mesencephalic graft transplantgroup are denoted by asterisks (P # 0.05).

536 C.E. SORTWELL ET AL.

grafted mesencephalic cell number was decreased (from500,000 to 280,000) DA neuron survival in the mixedstriatal co-graft condition increased from three-to-four-fold compared with mesencephalon grafted alone (Costan-tini et al., 1997).

Dense TH-immunoreactive patches in mixed striatalco-grafts were observed in the present study, confirmingprevious observations (Jaeger, 1986; Costantini et al.,1994). Both a contact-dependent phenomena as well asdiffusible substances have been implicated in the enhance-ment of mesencephalic DA neuron survival, whereas theeffects on nerve terminal differentiation are attributed todirect contact with striatal target neurons (Dong et al.,1993). These in vitro findings parallel what we observed inour present study. Separate co-grafts, in which diffusiblecommunication is maintained, but direct contact betweenstriatal and mesencephalic cells is abolished, producedincreased DA survival. However, mixed co-grafts, in whichcellular contact between the two tissue types occurs, bothexhibited increased DA neuron survival and containedzones of intensely TH immunoreactive patches within thegraft.

Whereas it was not surprising that the separate striatalco-grafts displayed significant behavioral recovery giventhat there is a reported correlation between DA neuronsurvival and behavioral recovery (Rioux et al., 1991), whatwas unexpected was the lack of behavioral improvement inthe mixed co-graft group. This may be related to a prefer-

ence of the grafted DA neurons in the mixed condition toinnervate the co-grafted striatal cells rather than the hoststriatum (DeBeaurepaire and Freed, 1987). In fact, TH1innervation within patches of mixed striatal co-grafts wasextremely dense, exceeding the fiber density of the intacthost striatum by 40%. However, enhanced reinnervation ofthe host striatum does not occur with the mixed co-graftswhen compared with solo grafts. Thus, the separate stria-tal co-graft approach seems optimal to enhance both DAneuron survival and reinnervation of the host striatum.

There is abundant evidence in vitro that striatal targettissue exerts tropic effects on the directionality and lengthof mesencephalic DA neurites. In the presence of striatalcells, mesencephalic cells show stimulated development(Prochiantz et al., 1979) and enhanced DA uptake, synthe-sis, and release (Prochiantz et al., 1979; Daguet et al.,1980). Additionally, mesencephalic neuronal processes arelengthened by co-culturing with striatal cells and solublestriatal extracts (Hemmendinger et al., 1981; Tomozawaand Appel, 1986). The degree of dopaminergic innervationto the denervated striatum and subsequent integrationinto the host striatal circuit also is considered crucial toestablishing graft efficacy (Bjorklund et al., 1987; Bjork-lund, 1992). Since the diffusion of DA is a restrictedphenomenon (Sendelbeck and Urquhart, 1985), tonic diffu-sion from grafted DA neurons can only replenish a limitedarea of the striatum. A common finding in rodents is thatDA neuron fiber outgrowth is limited to the immediatevicinity of the graft. In investigations in rodents where DAneuron fiber density and length is reported, the greatestlengths that fibers are seen penetrating is only 1–2 mminto the host caudate nucleus (Freed et al., 1980; Freed,1983; Stromberg et al., 1985; Mahalik et al., 1985). Further-more, the average DA terminal density in the graft-reinnervated striatum for a graft containing 1,000 DAneurons has been estimated to be around 40% of normal(Doucet et al., 1990; Manier et al., 1991). In the presentstudy, the separate striatal co-grafts produced signifi-cantly enhanced TH1 neurite outgrowth into the surround-ing striatum, particularly in host striatal regions inter-posed between the mesencephalic and striatal grafts,when compared with solo or mixed co-graft neurite out-growth. This separate co-graft-induced increase in reinner-vation density reached an average level of 50% of intactstriatum. These results confirm a tropic effect of theseparate striatal co-grafts on both the directionality anddensity of TH1 innervation, as suggested previously (Yureket al., 1990).

The only co-grafted cells used in the present study werestriatal cells; therefore whether or not other co-graftedtarget and nontarget cells can exert trophic effects is avalid question. Data generated in previous studies suggestthat, compared with other target and nontarget tissue,striatal target tissue exerts the maximal effect on bothtropic and trophic parameters in mesencephalic grafts andcultures. For example, maximal DAneuron axonal matura-tion and upregulation of DA transporter mRNA occurswhen mesencephalon is co-cultured with striatum, withsignificantly smaller effects elicited by target cortical cells,and no effect exerted by nontarget rostral tectum (Hem-mendinger et al., 1981; Perrone-Capano et al., 1996).Similar effects were reported on DA neuron survival whereagain striatal co-culturing induced maximal effects on DAneuron survival when compared with the lesser effects or

Fig. 7. Separate striatal co-grafts increase the dopaminergic rein-nervation of the adjacent host striatum. Fields (1003) within the hoststriatum immediately lateral to the transplant site were measured fordensity of TH1 reinnervation. Groups include sections from: mesence-phalic solo grafts (Solo Graft, n 5 5), mixed striatal-nigral co-grafts(Mixed Co-graft, n 5 5), separate striatal co-grafts (Sep Co-graft, n 55), lesioned striatum only (Lesioned, n 5 6), and unlesioned striatum(Unlesioned, n 5 6). Values represent the mean 6 SEM for each groupat 9 weeks following transplantation. All transplant groups and theunlesioned striatum possessed significantly denser TH1 striatalinnervation than lesioned striatum. Significant increases in TH1striatal density compared with both the solo mesencephalic grafttransplant group and the mixed co-graft group are denoted byasterisks (P # 0.01).

STRIATAL CO-GRAFTS AND DOPAMINE NEURON SURVIVAL 537

nonexistent effects of target cortex and nontarget cerebel-lum, hippocampus, olfactory bulb, and liver (Tomazawaand Appel, 1986; Dong et al., 1993). In rats, only nigraltissue and cells co-grafted with striatum and not othertissues including cerebellum, cortex, and spinal cord wereeffective in promoting extensive fiber outgrowth and is-lands of intense catecholamine fluorescence (Brundin etal., 1986; Jaeger, 1986; Dunnett et al., 1989). These manyexperiments provide evidence to support the concept thatthe effect observed in the present experiment on DAneuron survival and fiber outgrowth is specific for theco-grafted striatal target cells.

Separate and mixed striatal co-grafts had equivalenteffects on enhancing mesencephalic DA neuron survival.Therefore it is likely that the striatal co-graft serves as asource of diffusible neurotrophic substances that are notprovided by the adult host striatum, or are provided inincreased amounts by embryonic striatum. Several knowngrowth factors, such as basic fibroblast growth factor(bFGF; Knusel et al., 1990), epidermal growth factor (EGF;Casper et al., 1991), insulin-like growth factor 1 (IGF-1;Knusel et al., 1990), brain-derived neurotrophic factor(BDNF; Beck et al., 1993), neurotrophin-3 (Hyman et al.,1993), neurotrophin-4/5 (Hynes et al., 1993), and glial cellline-derived growth factor (GDNF; Lin et al., 1993) havebeen demonstrated to promote survival of DA neurons tovarying degrees, acting either directly or indirectly. Sev-eral known trophic factors, such as BDNF (Spenger et al.,1995), neurotrophin-3 and neurotrophin-4/5 (Hyman etal., 1994), and GDNF (Lin et al., 1993) have been demon-strated to promote survival of DA neurons in vitro tovarying degrees. However, to date, only GDNF has beenlocalized specifically to striatal target areas during devel-opment (Schaar et al., 1993; Stromberg et al., 1993;Choi-Lundberg and Bohn, 1995). In the adult striatum,high to moderate expression of GDNF, platelet-derivedgrowth factor (PDGF), and transforming growth factor(TGF) has been detected (Sasahara et al., 1991; Seroogy etal., 1991; Unsicker et al., 1991; Yeh et al., 1991; Springer etal., 1994; Blum and Weickert, 1995; Trupp et al., 1997).Extremely low levels of BDNF are detected in the adultstriatum with immunocytochemistry; however mRNA lev-els indicate that BDNF is not present in adult striatumunless seizures are induced (Schmidt-Kastner et al., 1996)or dopaminergic stimulation occurs (Okazawa et al., 1992).This suggests that in the adult, levels of BDNF are notconstant and that production of BDNF may increase toplay a role in a lesioned or injury-induced state. It also ispossible that undefined striatal-derived trophic factorsmay be released by the striatal co-graft, such as thoseinvolved in the trophic effects of striatal extracts (Dal Tosoet al., 1988) and striatal O-2A progenitor cells (Takeshimaet al., 1994).

Since multiple trophic factors can influence nigral DAneurons, striatal co-grafts may provide grafted DA neu-rons with a continuous source of endogenous factors attheir appropriate developmental timepoints. Studies ofinfusion of a variety of growth factors in conjunctionwith neural grafts reveal enhanced graft viability both inoculo and in the denervated striatum (Giacobini et al.,1990, 1991; Steinbusch et al., 1990; Stromberg et al., 1993;Takayama et al.,1995; Rosenblad et al., 1996; Wang et al.,1996). As it may prove difficult to infuse appropriately theidentical cascade of trophic factors that are provided by the

striatal co-graft, the co-graft approach may prove compa-rable or even superior in exerting trophic/tropic effects.

The mechanism by which co-cultured and co-graftedstriatal cells enhance the survival of DA neurons may beby supplying trophic factors that protect DA neurons fromprogrammed cell death, i.e., apoptosis. These factors maybe unavailable in the adult host striatum. The ‘‘neuro-trophic factor hypothesis’’ suggests that survival of devel-oping neurons depends on the secretion of specific factorsfrom target cells that the neurons innervate. These trophicsignals appear to exert their supportive effect by suppress-ing an intrinsic cell ‘‘suicide’’ program (Raff et al., 1993),i.e., an apoptotic program that engages when such signalsare absent. In cultures of sympathetic neurons, nervegrowth factor (NGF) deprivation induces apoptosis,whereas NGF-exposed neurons are protected (Deckwerthand Johnson, 1993). Exposure of mesencephalic cultures toGDNF reduces the rate of apoptotic cell death (Clarkson etal., 1995). In the nigrostriatal system, lesions of thestriatum during development result in apoptosis in thesubstantia nigra (Macaya et al., 1994). Until recently, itwas assumed that cell death in mesencephalic graftsprimarily was necrotic, i.e., the result of inadequate vascu-larization or trauma to the graft during tissue dissection ortransplantation. However, apoptosis has been suggested tobe a normal part of graft development, particularly atearly times following grafting (Mahalik et al., 1994;Sortwell et al., 1997). These findings suggest that graftsmay be under the guidance of the same genetic mecha-nisms that function in naturally developing neural sys-tems and may offer insights into how striatal co-graftsenhance survival of grafted mesencephalic neurons throughthe inhibition of apoptosis.

The development of strategies that increase grafted DAneuron survival and promote innervation of the hoststriatum are critical to the success of neural transplanta-tion for PD. Currently, the survival of DA neurons inhuman grafts is estimated to be about 5–10% (Olanow etal., 1996) and although neurite extension from an averagegraft can extend through a large cross-sectional areawithin the caudate, the density of this reinnervationapproximates only 40% of that seen in the intact striatum(Doucet et al., 1990; Manier et al., 1991). These shortcom-ings suggest that the use of multiple donors and multipleplacement sites may be necessary for enhanced and sus-tained therapeutic benefits. The use of multiple donorsmay be impractical and lead to other problems, such asimmunological rejection. An alternate strategy is to in-crease the survival and neurite outgrowth of graftedDA neurons. Taken collectively, our data suggest that1) embryonic striatal co-grafts exert a trophic effect on DAneurons by significantly augmenting cell survival, 2) stria-tal co-grafts placed separately at short distances also exerta tropic effect that significantly enhances dopaminergicinnervation of the lesioned striatum, and 3) enhancedneurite outgrowth and augmented DA neuron survivalmay participate in enhanced restoration of neural cir-cuitry necessary for behavioral recovery from parkinson-ism.

ACKNOWLEDGMENTS

We are grateful for the excellent technical advice andassistance of Barbara Blanchard and Brian Daley. Thiswork was supported by NIH grants F32 NS09646–01

538 C.E. SORTWELL ET AL.

(to C.E.S.), RSA K05-MH 00643(to D.E.R.), and AG10851(T.J.C.).

LITERATURE CITED

Abercrombie, M. (1946) Estimation of nuclear populations from microtomesections. Anat. Rec. 94:239–247.

Ahlskog, J.E. (1993) Cerebral transplantation for Parkinson’s disease:Current progress and future prospects. Mayo Clin. Proc. 68:578–591.

Beck, K.D., B. Knusel, and F. Hefti (1993) The nature of the trophic action ofbrain-derived neurotrophic factor, des(1-3)-insulin-like growth factor-1,and basic fibroblast growth factor on mesencephalic dopaminergicneurons developing in culture. Neurosci. 52:855–866.

Bjorklund, A. (1992) Dopaminergic transplants in experimental parkinson-ism: Cellular mechanisms of graft-induced functional recovery. Curr.Opin. Neurobiol. 2:683–689.

Bjorklund, A., S.B. Dunnett, U. Stenevi, M.E. Lewis, and S.D. Iversen(1980) Innervation of the denervated striatum by substantia nigratransplants: Functional consequences as revealed by pharmacologicaland sensorimotor testing. Brain Res. 199:307–333.

Bjorklund, A., O. Lindvall, O. Isacson, P. Brundin, K. Wictorin, R.E.Strecker, D.J. Clarke, and S.B. Dunnett (1987) Mechanisms of action ofintracerebral neural implants: Studies on nigral and striatal grafts tothe lesioned striatum. TINS 10:509–516.

Blum, M. and C.S. Weickert (1995) GDNF mRNA expression in normalpostnatal development, aging and in weaver mutant mice. Neurobiol.Aging 16:925–929.

Brundin, P., O. Isacson, and A. Bjorklund (1985) Monitoring of cell viabilityin suspensions of embryonic tissue and its use as a criterion forintracerebral graft survival. Brain Res. 331:251–259.

Brundin, P., O. Isacson, F.H. Gage, and A. Bjorklund (1986) Intrastriatalgrafting of dopamine-containing neuronal cell suspensions: Effects ofmixing with target or non-target cells. Dev. Brain Res. 24:77–84.

Carvey, P.M., D.H. Lin, C.J. Faselis, J.K. Noterman, and Z.D. Ling (1996)Loss of striatal innervation increases striatal trophic activity directedat DA neurons in culture. Exp. Neurol. 140:184–197.

Casper, D., C. Mytilineou, and M. Blum (1991) EGF enhances the survivivalof dopamine neurons in rat embryonic mesencephalon primary culture.J. Neurosci. 30:372–381.

Choi-Lundberg, D.L. and M.C. Bohn (1995) Ontogeny and distribution ofglial cell line-derived neurotrophic factor (GDNF) mRNA in rat. Dev.Brain Res. 85:80–88.

Clarkson, E.D., W.M. Zawada, and C.R. Freed (1995) GDNF reducesapoptosis in dopaminergic neurons in vitro. Neuroreport 7:145–149.

Costantini, L.C., B.M. Vozza, and A.M. Snyder-Keller (1994) Enhancedefficacy of nigral-striatal cotransplants in bilaterally dopamine-depleted rats: An anatomical and behavioral analysis. Exp. Neurol.127:219–231.

Costantini, L.C. and A.M. Snyder-Keller (1997) Co-transplantation of fetallateral ganglionic eminence and ventral mesencephalon can augmentfunction and development of intrastriatal transplants. Exp. Neurol.145:214–227.

Daguet, M.-C., U. Di Porzio, A. Prochiantz, A. Kato, and J. Glowinski (1980)Release of dopamine from dissociated mesencephalic dopaminergicneurons in primary cultures in absence or presence of striatal targetcells. Brain Res. 191:564–568.

Dal Toso, R., O. Giorgi, C. Soranzo, G. Kirschner, G. Ferrari, M. Favaron, D.Benvegnu, D. Presti, S. Vicini, G. Toffano, G.F. Azzone, and A. Leon(1988) Development and survival of neurons in dissociated fetal mesen-cephalic serum-free cell cultures: I. Effects of cell density and of anadult mammalian striatal-derived neuronotrophic factor (SDNF). J.Neurosci. 8:733–745.

Deacon, T.W., P. Pakzaban, and O. Isacson (1994) The lateral ganglioniceminence is the origin of cells committed to striatal phenotypes: Neuraltransplantion and developmental evidence. Brain Res. 668:211–219.

DeBeaurepaire, R. and W.J. Freed (1987) Embryonic substantia nigragrafts innervate embryonic striatal co-grafts in preference to maturehost striatum. Exp. Neurol. 95:448–454.

Deckwerth, T.L. and E.M. Johnson, Jr. (1993) Temporal analysis of eventsassociated with programmed cell death (apoptosis) of sympatheticneurons deprived of nerve growth factor. J. Cell Biol. 123:1207–1222.

Dong, J.F., A. Detta, M.H.M. Bakker, and E.R. Hitchcock (1993) Directinteraction with target-derived glia enhances survival but not differen-tiation of human fetal mesencephalic dopaminergic neurons. Neurosci-ence 56:53–60.

Doucet, G., P. Brundin, I. Descarries, and A. Bjorklund (1990) Effect of priordenervation on survival and fiber outgrowth from intrastriatal fetalmesencephalic grafts. Eur. J. Neurosci. 2:279–290.

Dunnett, S.B., A. Bjorklund, U. Stenevi, and S.D. Iversen (1981) Behavioralrecovery following transplantation of substantia nigra in rats subjectedto 6-OHDA lesions of the nigrostriatal pathway. I. Unilateral lesions.Brain Res. 215:147–161.

Dunnet, S.B., A. Bjorklund, R.H. Schmidt, U. Stenevi, and S.D. Iversen(1983) Intracerebral grafting of neuronal cell suspensions IV. Behav-ioral recovery in rats with unilateral 6-OHDA lesions following implan-tation of nigral cell suspensions in different forebrain sites. ActaPhysiol. Scand. Suppl. 522:29–37.

Dunnett, S.B., D.C. Rogers, and S.J. Richards (1989) Nigrostriatal recon-struction after 6-OHDA lesions in rats: Combination of dopamine-richnigral grafts and nigrostriatal ‘‘bridge’’ grafts. Exp. Brain Res. 75:523–535.

Finger, S. and S.B. Dunnett (1989) Nimodipine enhances growth andvascularization of neural grafts. Exp. Neurol. 104:1–9.

Freed, W.J. (1983) Functional brain tissue transplantation: Reversal oflesion-induced rotation by intraventricular substantia nigra and adre-nal medulla grafts, with a note on intracranial retinal grafts. Biol.Psychiatry 18:1205–1267.

Freed, W.J., M.J. Perlow, F. Karoum, A. Seiger, L. Olson, B.J. Hoffer, andR.J. Wyatt (1980) Restoration of dopaminergic function by grafting offetal rat substantia nigra to the caudate nucleus: Long-term behavioral,biochemical and histochemical studies. Ann. Neurol. 8:510–519.

Freeman, T.B., C.W. Olanow, R.A. Hauser, G.M. Nauert, D.A. Smith, C.V.Borlongan, P.R. Sanberg, D.A. Holt, J.H. Kordower, and A.B. Vinger-hoet (1995) Bilateral fetal nigral transplantation into the postcommis-ural putamen in Parkinson’s disease. Ann. Neurol. 38:379–388.

Giacobini, M.M.J., L. Olson, B.J. Hoffer, and V.R. Sara (1990) TruncatedIGF-1 exerts trophic effects on fetal brain tissue grafts. Exp. Neurol.108:33–37.

Giacobini, M.M.J., B.J. Hoffer, G. Zerbe, and L.Olson (1991) Acidic andbasic fibroblast growth factors augment growth of fetal brain tissuegrafts. Exp. Brain Res. 86:73–81.

Hemmendinger, L.M., B.B. Garber, P.C. Hoffman, and A. Heller (1981)Target neuron-specific process formation by embryonic mesencephalicdopamine neurons in vitro. Proc. Natl. Acad. Sci. USA 78:1264–1268.

Hoffman, P.C., L.M. Hemmendinger, C. Kotake, and A. Heller (1983)Enhanced dopamine cell survival in reaggregates containing telence-phalic targets. Brain Res. 274:275–281.

Hornykiewicz, O. (1988) Neurochemical pathology and the etiology ofParkinson’s disease: Basic facts and hypothetical possibilities. Mt.Sinai J. Med.55:11–20.

Hyman, C., M. Juhasz, C. Jackson, P. Wright, N.Y. Ip, and R.M. Lindsay(1994) Overlapping and distinct actions of the neurotrophins BDNF,NT-3 and NT-4/5 on cultured dopaminergic and GABAergic neurons ofthe ventral mesencephalon . J. Neurosci. 4:335–3477.

Hynes, M.L., L. Berkemeier, K. Poulsen, F. Hefti, and A. Rosenthal (1993)Neurotrophin-4/5 promotes survival of dopaminergic neurons, and ispresent in embryonic striatum. Soc. Neuro. Abst. 19:653.

Jaeger, C.B. (1986) Axon terminal clustering in nigrostriatal double grafts.Dev. Brain Res. 24:309–314.

Knusel, B., P. Mishel, J.S. Schwaber, and F. Hefti (1990) Selective andnonselective stimulation of central cholinergic and dopaminergic devel-opment in vitro by nerve growth factor, basic fibroblast growth factor,epidermal growth factor, insulin and the insulin-like growth factors Iand II. J. Neurosci. 10:558–570.

Lin, L-FH., D.H. Doherty, J.D. Lile, S. Bektesh, and F. Collins (1993)GDNF: A glial cell line-derived neurotrophic factor for midbrain dopa-minergic neurons. Science 260:1130–1132.

Macaya, A., F. Munell, R.M. Gubits, and R.E. Burke (1994) Apoptosis insubstantia nigra following developmental striatal exitotoxic injury.Proc. Natl. Acad. Sci. USA 91:8117–8121.

Mahalik, T.J., T.E. Finger, I. Stromberg, and L. Olson (1985) Substantianigra transplants into denervated striatum of the rat: Ultrastructure ofgraft and host interconnections. J. Comp. Neurol. 240:60–70.

Mahalik, T.J., W.E. Harin, G.H. Clinton, and G.P.Owens (1994) Pro-grammed cell death in developing grafts of fetal substantia nigra. Exp.Neurol. 129:27–36.

Manier, M., D.N. Abrous, C. Feurerstein, M. LeMoal, and J.P. Herman(1991) Lesion of the nigrostriatal dopaminergic pathway in adult ratsand reversal following the implantation of embryonic dopaminergicneurons: A quantitative immunohistochemical analysis. Neuroscience42:427–439.

STRIATAL CO-GRAFTS AND DOPAMINE NEURON SURVIVAL 539

Niijima, K., M. Araki, M. Ogawa, I. Nagatsu, F. Sato, H. Kimura, and M.Yoshida (1990) Enhanced survival of cultured dopamine neurons bytreatment with soluble extracts from chemically deafferentiated stria-tum of adult rat brain. Brain Res. 528:151–154.

Nikkhah, G., M.G. Cunningham, A. Jodicke, U. Knappe, and A. Bjorklund(1994) Improved graft survival and striatal reinnervation by microtrans-plantation of fetal nigral cell suspensions in the rat Parkinson model.Brain Res. 633:133–143.

Okazawa, H., M. Murata, M. Watanabe, M. Kamei, and I. Kanazawa (1992)Dopaminergic stimulation up-regulates the in vivo expression of brain-derived neurotrophic factor (BDNF) in the striatum. FEBS Lett.313:138–142.

Olanow, C.W., J.H. Kordower, and T.B. Freeman (1996) Fetal nigraltransplantation as a therapy for Parkinson’s disease. TINS 19:102–109.

Olson, L., A. Seiger, B. Hoffer, and D. Taylor (1979) Isolated catecholaminer-gic projections from substantia nigra and locus coeruleus to caudate,hippocampus and cerebral cortex formed by intraocular sequentialdouble brain grafts. Exp. Brain Res. 35:47–67.

Pakzaban, P., T.W. Deacon, L.H. Burns, and O. Isacson (1993) Increasedproportion of acetylcholinesterase-rich zones and improved morphologi-cal integration in host striatum of fetal grafts derived from lateral butnot medial ganglionic eminence. Exp. Brain Res. 97:13–22.

Perrone-Capano, C., A. Tino, G. Amadoro, R. Pernas-Alonso, and U. DiPorzio (1996) Dopamine transporter gene expression in rat mesence-phalic dopaminergic neurons is increased by direct interaction withtarget striatal cells in vitro. Mol. Brain Res. 39:160–166.

Prochiantz, A., U. Di Porzio, A. Kato, B. Berger, and J. Glowinski (1979) Invitro maturation of mesencephalic dopaminergic neurons from mouseembryos is enhanced in the presence of their striatal target cells. Proc.Natl. Acad. Sci. USA 76:5387–5391.

Raff, M.C., B.A. Barres, J.F. Burne, H.S. Coles, Y. Ishizaki, and M.D.Jacobson (1993) Programmed cell death and the control of cell survival:Lessons from the nervous system. Science 262:695–700.

Rioux, L., D.P. Gaudin, L.K. Bui, L. Gregoire, T. DiPaolo, and P.J. Bedard(1991) Correlation of functional recovery after a 6-hydroxydopaminelesion with survival of grafted fetal neurons and release of dopamine inthe striatum of the rat. Neuroscience 40:123–131.

Rosenblad, C., A. Martinez-Serrano, and A. Bjorklund (1996) Glial cellline-derived neurotrophic factor increases survival, growth and func-tion of intrastriatal fetal nigral dopaminergic grafts. Neuroscience75:979–985.

Sasahara M., J.W.U. Fries, E.W. Raines, A.M. Gown, L.E. Westrum, M.P.Frosch, D.T. Bonthron, R. Ross, and T. Collins (1991) PDGF B-chain inneurons of the central nervous system, posterior pituitary, and in atransgenic model. Cell 64:217–227.

Schaar, D.G., B.-A. Sieber, C.F. Dreyfus, and I.B. Black (1993) Regional andcell-specific expression of GDNF in rat brain. Exp. Neurol. 124:368–371.

Schmidt-Kastner, R., C. Wetmore, and L. Olson (1996) Comparative studyof brain-derived neurotrophic factor messenger RNA and protein at thecellular level suggests multiple roles in hippocampus, striatum andcortex. Neuroscience 74:161–183.

Sendelbeck, S.L. and J. Urquhart (1985) Spatial distribution of dopamine,methotrexate and antipyrine during continuous intracerebral microper-fusion. Brain Res. 328:251–258.

Seroogy, K.B., V.K.M. Han, and D.C. Lee (1991) Regional expression oftransforming growth factor-alpha mRNA in the rat central nervoussystem. Neurosci. Lett. 125:241–245.

Sladek Jr., J.R., T.J. Collier, J.D. Elsworth, J.R. Taylor, R.H. Roth, and D.R.Redmond, Jr. (1993) Can graft-derived neurotrophic activity be used todirect axonal outgrowth of grafted dopamine neurons for circuit recon-struction in primates? Exp. Neurol. 124:134–139.

Sortwell, C.E., T.J. Collier, B.C. Blanchard, and J.R. Sladek, Jr. (1996)Analysis of tropic and trophic effects on mesencephalic grafts to thelesioned rat striatum by cografted striatal cells. Soc. Neurosci. Abstr.22:1493.

Sortwell, C.E., T.J. Collier, and J.R. Sladek, Jr. (1997) Cografted embryonicstriatum enhances grafted mesencephalic neuron viability: Evidencefor the inhibition of apoptosis. In: Proceedings of the 6th InternationalNeural Transplantation Meeting, p. 22.

Spenger, C., C. Hyman, L. Studer, M. Egli, L. Evtouchenko, C. Jackson, A.Dahl-Jorgensen, R. Lindsay, and R.W. Seiler (1995) Effects of BDNF ondopaminergic, serotonergic, and GABAergic neurons in cultures ofhuman fetal ventral mesencephalon. Exp. Neurol. 133:50–63.

Springer, J.E., X. Mu, L.W. Bergmann, and J.Q. Trojanowski (1994)Expression of GDNF mRNA in rat and human nervous tissue. Exp.Neurol. 127:167–170.

Steinbusch, H.W.M., R.J. Vermeulen, and J.A.D.M. Tonnaer (1990) Basicfibroblast growth factor enhances survival and sprouting of fetaldopaminergic cells implanted in the denervated rat striatum. Prog.Brain Res. 82:81–86.

Stromberg, I., S. Johnson, B. Hoffer, and L. Olson (1985) Reinnervation ofdopamine-denervated striatum by substantia nigra transplants: Immu-nohistochemical and electrophysiological correlates. Neuroscience 14:981–990.

Stromberg, I., L. Bjorklund, M. Johansson, A. Tomac, F. Collins, L. Olson, B.Hoffer, and C. Humpel (1993) Glial cell line-derived neurotrophic factoris expressed in the developing but not adult striatum and stimulatesdeveloping dopamine neurons in vivo. Exp. Neurol. 124:401–412.

Takayama, H., J. Ray, H.K. Raymon, A. Baird, J. Hogg, L.J. Fisher, and F.H.Gage (1995) Basic fibroblast growth factor increases dopaminergic graftsurvival and function in a rat model of Parkinson’s disease. NatureMed. 1:53–58.

Takeshima, T., J.M. Johnston, and J.W. Commissiong (1994) Oligodendro-cyte-type-2 astrocyte (O-2A) progenitors increase the survival of ratmesencephalic, dopaminergic neurons from death induced by serumdeprivation. Neurosci. Lett. 166:178–182.

Taylor, J.R., J.D. Elsworth, R.H. Roth, T.J. Collier, J.R. Sladek, Jr., and D.E.Redmond, Jr. (1990) Improvements in MPTP-induced object retrievaldeficits and behavioral deficits after fetal nigral grafting in monkeys.Prog. Brain Res. 82:543–559.

Taylor, J.R., J.D. Elsworth, R.H. Roth, J.R. Sladek, Jr., T.J.Collier, and D.E.Redmond Jr. (1991) Grafting of fetal substantia nigra to striatumreverses behavioral deficits induced by MPTP in primates: A compari-son with other types of grafts as controls. Exp. Brain Res. 85:335–348.

Tomozawa, Y. and S.H. Appel (1986) Soluble striatal extracts enhancedevelopment of mesencephalic dopaminergic neurons in vitro. BrainRes. 399:111–124.

Trupp, M., N. Belluardo, H. Funakoshi, and C.F. Ibanez (1997) Complemen-tary and overlapping expression of glial cell line-derived neurotrophicfactor (GDNF), c-ret proto-oncogene, and GDNF receptor-alpha indi-cates multiple mechanisms of trophic actions in the adult rat CNS. J.Neurosci. 17:3554–3567.

Unsicker, K., K.C. Flanders, D.S. Cissel, R. Lafyatis, and M.B. Sporn (1991)Transforming growth factor beta isoforms in the adult rat central andperipheral nervous sytem. Neuroscience 44:613–625.

Wang, Y., L.T. Tien, P.A. Lapchak, and B.J. Hoffer (1996) GDNF triggersfiber outgrowth of fetal ventral mesencephalic grafts from nigra tostriatum in 6-OHDA-lesioned rats. Cell Tissue Res. 286:225–233.

Yeh, H.-J., K.G. Ruit, Y.-X. Wang, W.C. Parks, W.D. Snider, and T.F. Deuel(1991) PDGF A-chain gene is expressed by mammalian neurons duringdevelopment and in maturity. Cell 64:209–216.

Yurek, D.M., T.J. Collier, and J.R. Sladek, Jr. (1990) Embryonic mesence-phalic and striatal co-grafts: Development of grafted dopamine neuronsand functional recovery. Exp. Neurol. 109:191–199.

540 C.E. SORTWELL ET AL.