Influence of BDNF on the expression of the dopaminergic phenotype of tissue used for brain transplants

Developmental Brain Research 100 Ž1997. 43–51

Research report

Influence of BDNF on the expression of the dopaminergic phenotype of tissue used for brain transplants Jiawei Zhou a , Henry F. Bradford a

a,)

, Gerald M. Stern

b

Department of Biochemistry, Imperial College of Science, Technology and Medicine, London SW7 2AY, UK b Department of Clinical Neurology, UniÕersity College, School of Medicine, London W1N 8AA, UK Accepted 30 December 1996

Abstract Brain-derived neurotrophic factor ŽBDNF. has previously been shown by this laboratory among others to promote survival and differentiation of central dopaminergic neurons and to stimulate expression of the dopaminergic phenotype in fetal cerebrocortex in vitro. We have examined the effect of BDNF antibody on nigral dopaminergic neurons in vivo and in vitro. It reduced the survival of rat fetal dopaminergic neurons in culture Žup to 40% died.. The BDNF antibody also caused ipsilateral rotation after a single in vivo intranigral injection in the adult rats. Pre-treatment of fetal nigral neurons with BDNF improved the performance of dopaminergic cells in fetal nigral transplants based on surviving THq cells numbers. Thus, parkinsonian rats receiving fetal nigral cells treated with BDNF showed a significantly greater reduction of turning over the 3 weeks following transplantation, compared with the rats receiving untreated nigral transplants. However, the average number of tyrosine hydroxylase ŽTH.-positive neurons in the grafts of rats receiving fetal nigral cells treated with BDNF was 211 " 35 which was only about 20% of the cell number Ž1012 " 223, mean " S.E.M.. found in those receiving untreated nigral transplants. These results suggest that pretreatment of nigral dopaminergic neurons with BDNF may improve their functional performance, but not their survival in transplants. The ability of artificially induced cerebrocortical ‘dopaminergic’ cells to ameliorate behavioral asymmetry of Parkinsonian rats was assessed. A proportion Ž1.0% maximum. of the THq neurons in these transplants survived in the host brain and were likely to be responsible for the prominent reduction in rotation scores observed to occur 6 weeks after implantation. Thus, the combined treatment of fetal cerebral cortex with BDNF and dopamine created long-lived TH-expressing neuronal populations which were very effective in alleviating the rat parkinsonian model, and thus may be suitable for use in transplantation in treating human Parkinson’s disease. Keywords: Nerve growth factor; Brain-derived neurotrophic factor; Neural transplantation; Parkinson’s disease

1. Introduction Neural transplantation has brought promising prospects for improving treatment of neurodegenerative diseases in general. In spite of the great progress made, there are still many central problems to be resolved. For example, the low survival rate of fetal neurons after grafting. It is estimated that up to 90–95% of neurons fail to survive after implantation w3x. The reasons for this are poorly understood as yet. Lack of sufficient neurotrophic support for these cells is suggested to be a one of the reasons. It may be possible to increase the survival rate of grafted cells by administering certain neurotrophic substances prior to transplantation or after transplantation. This approach is supported by the finding that pretreatment of neural graft )

Corresponding author. Fax: q44 Ž171. 225 0960.

tissue by brief incubation with basic fibroblast growth factor ŽbFGF., or repeated intrastriatal infusion of bFGF can increase the survival of dopaminergic neurons in fetal nigral grafts w13x. It is suggested that mesencephalic tissue from as many as 10–15 fetuses may be necessary for fully efficient transplantation into striata of a single patient if bilateral grafting proves necessary w2,3x. To get so many fetal brains at the same time proves to be very difficult practically. These difficulties may be overcome with the use of primary cultures of human brain tissue together with their cryopreservation in a ‘brain bank’. Studies in this laboratory and elsewhere have shown that cultured human fetal mesencephalic tissue can relieve the Parkinsonian syndrome in rats, but no such recovery was observed following implantation of human fetal cortical cultures w22x. This suggests that cultures of this sort would provide useful

0165-3806r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 3 8 0 6 Ž 9 7 . 0 0 0 1 9 - 9

44

J. Zhou et al.r DeÕelopmental Brain Research 100 (1997) 43–51

neural tissue for transplantation therapy in Parkinson’s disease. BDNF has been shown to promote in vitro survival of rat and human embryonic dopaminergic neurons w11,24x, and also to protect these neurons against MPTP toxicity w10x. Co-treatment of fetal cortical neurons with dopamine ŽDA. and BDNF induces the full dopaminergic phenotype, including tyrosine hydroxylase ŽTH. induction, DA biosynthesis and high-affinity DA uptake w23,25x. Others have shown TH induction in fetal striatal neurons w4,5x. In the present study, BDNF was tested to see whether: Ža. pretreatment of fetal nigral cells in cultures with BDNF could improve their performance following transplantation into 6-hydroxydopamine Ž6-OHDA.-lesioned rats; Žb. dopaminergic tissue artificially induced in fetal cerebral cortex could cure experimental Parkinsonism, and Žc. administration of an antibody to BDNF in vitro and in vivo could disrupt the dopaminergic system.

antibody was affinity-purified rabbit anti-BDNF antibody ŽAmgen, USA, w15x., which was diluted in 0.1 M carbonate-bicarbonate buffer ŽpH 9.6. at 1 m grml. Following incubation with 1% BSA, 50 m l of culture medium or standards were added and the plates were kept overnight at 48C. The detecting antibody used was turkey anti-BDNF antiserum Ž1:1000 dilution.. After incubation with biotinylated rabbit anti-turkey IgG Ž1 m grml. and streptavidinalkaline phosphatase, the alkaline phosphatase activity was measured using p-nitrophenyl phosphate ŽpNPP, Sigma 104 w . as the substrate, dissolved in 0.1 M diethanolamine buffer ŽpH 9.6, BDH, UK. containing 1 mM MgCl 2 . The absorbance was observed at a wavelength of 405 nm with an EIA Reader. The BDNF content of each sample Žduplicates or triplicates. was calculated from the linear portion of the standard curve and expressed in ngrml culture medium. Controls were identical to the standard but without addition of primary antibody. 2.3. Application of BDNF antibody

2. Materials and methods 2.1. Animals and preparation of dissociated cell cultures The Sprague-Dawley rats, provided by the Animal House at Imperial College, London, were caged in groups of three with free access to food and water. The animals were kept in a temperature controlled environment Ž218C. on a 12 h lightrdark cycle Žlight period 07.00–19.00 h.. Fetal cerebrocortices and ventral mesencephalons were obtained at day 12 and 14 ŽE12–14, E0 s day of vaginal plug. of gestation, immediately after the rats were sacrificed by decapitation. The rat embryos were stored in ice-cold Hank’s balanced saline solution ŽCa2q and Mg 2q free, HBSS, Gibco. until dissection. After removal of meninges, the embryos were dissected out in HBSS and cut into 0.5 mm sections. Following washing, the tissue was incubated in HBSS containing 1 mgrml trypsin and 0.5 mgrml DNase at 378C for 10 min. Tissue was then sequentially washed in HBSS containing DNase. Rat cells were seeded into 35 mm plastic Petri dishes ŽFalcon. or 16 mm multiwell plates ŽNunc, Denmark. precoated with poly-L-lysine Ž5 m grml. as described by Walters et al. w22x. Cell viability in suspensions of dissociated cells was determined by the ability of viable neurons to exclude the dye trypan blue. The cells were plated at 1 = 10 5 cellsrcm2 in Dulbecco’s modified Eagle’s medium ŽDMEM, Gibco, UK. supplemented with N2 Ž1:100., B27 Ž1:100, Gibco. and 50 mgrml gentamycin. Cells were incubated at 378C in a 95% airr5% CO 2 humidified atmosphere and the medium was changed twice a week.

2.3.1. Nature and source of BDNF antibody Rabbit and turkey antibodies were raised against human recombinant BDNF and generously supplied to us as a gift by Dr. James Miller and colleagues of Amgen, CA, USA. The rabbit antiserum was affinity-purified w15x. The turkey antiserum to BDNF and the rabbit BDNF antibody was tested in BDNF EIA and by Western blot. They did not react with other neurotrophins w26x. 2.3.2. Application of rabbit BDNF antibody to cultures of dopaminergic neurons Specific rabbit anti-BDNF antibody was added to rat mesencephalic cultures to neutralize BDNF present in the culture. Sodium azide, a preservative in the antibody, was removed by dialysis with a Microdialyser ŽSystem 100, Pierce. in 0.1 M PBS before the antibody was used. BDNF antibody Žfinal concentration: 1:100 s 8 m grml. was added to the cultures three times a week, i.e. days 0, 3, 5. 2.3.3. Intranigral in ÕiÕo injection of BDNF antibody To determine whether BDNF antibody influences the function of mesencephalic dopaminergic neurons in vivo, ten microliters of turkey anti-BDNF serum was injected into one side of the substantia nigra. The coordinates with tooth bar set at y3.3 mm were AP: y5.4 mm, L: 1.5 mm; V: 8 mm w19x. Two weeks after injection, amphetamine Ž5 mgrkg. was administered to test for locomotor asymmetry. The test was performed every 2–3 weeks thereafter using an automated rotometer apparatus devised by Dr S.B. Dunnett of Cambridge University, U.K. and constructed in his university workshop.

2.2. BDNF enzyme immunoassay (EIA)

2.4. Lesion surgery with 6-OHDA and transplantation

The measurement of BDNF using two-site sandwich EIA was described previously w26x. Briefly, the capture

Unilateral lesion of the nigrostriatal pathway Žright side. in adult female Sprague-Dawley rats Žbody weight 180–200

J. Zhou et al.r DeÕelopmental Brain Research 100 (1997) 43–51

g. was performed using 6-OHDA as described previously w21x. Amphetamine Ž5 mgrkg. was administered to test for locomotor asymmetry. This was tested every 3–5 weeks after grafting using an automated and computerized rotometer apparatus. Less than six to seven turnsrmin are internationally accepted to indicate ‘recovery’ or ‘normality’ w6,21x. Transplants were prepared as described previously w22x with a few modifications. Ventral mesencephalons and cortices were dissected from rat fetuses at E14 and E12 respectively and cultured in 35 mm Petri dishes separately with serum-containing medium following dissociation at a cell density of 10 6 cells per dish. 2.4.1. Fetal nigral grafts Four to five weeks after the 6-OHDA lesion, the animals, whose behavioral asymmetry was previously assessed and showed a rotation rate in excess of 6 turns per min after exposure to amphetamine, were divided into 5 groups. Each group consisted of 5 ; 7 animals Žsee Results ŽSection 3. for details.. Group SNB received nigral cell transplants pretreated with BDNF Ž50 ngrml.. Group SNC, as a control group, received nigral cell transplants without BDNF pretreatment. Approximately 5 = 10 5 cultured fetal mesencephalic cells were stereotaxically injected in 6 = 2 ml Žsee below. into the rat striatum, after harvesting. At an average of 2.0% THq cells w23x this indicates the presence of 10,000 THq neurons per graft w16x. To investigate the effects of BDNF treatment on cultured mesencephalic neurons, and their survival after grafting, one group of cultures was grown in the presence of BDNF Ž50 ngrml., that was added at day 0 and day 2. At day 4 after plating, cultures were gently harvested in HBSS ŽCa2q- and Mg 2q-free. containing DNase Ž0.1 mgrml. and EDTA Ž10 mM. using a rubber policeman, pooled in Eppendorf tubes and gently centrifuged Ž100 g, 5 min.. Most of the supernatant was discarded. The pellets were resuspended, and BDNF was then added into the cell suspension derived from the BDNF-treated cultures, to make a final concentration of 50 ngrml. The cell suspension Žcell concentration, 10 5 cells per microliter. was kept on ice until surgery was finished. 2.4.2. Fetal cerebral cortex grafts For grafting of the cortical cells, half the total number of cortical cell dishes were treated with BDNF Ž50 ngrml. and DA Ž10 m M. at day 1 and 3 after plating. All treated and untreated cells were then harvested as above. The cells were then resuspended to a concentration of 5 = 10 4 cells per microliter containing BDNF Ž50 ngrml. and DA Ž10 m M., DNase and EDTA were added. The cells suspension was kept on ice throughout the surgery. No addition of BDNF and DA was made to the cell suspension of the untreated group. Grafts Žnigral or cortical. were implanted by stereotaxic

45

microinjection of 2.0 m l of the nigral or cortical cell suspension via a 100-m l Hamilton syringe into each of 6 sites in the caudate-putamen ipsilateral to the nigrostriatal lesions. Thus, approximately 6 = 10 5 fetal cortical cells were injected. This would have included Žat an average of 6% w23x of the population., a total of approximately 3 = 10 4 THq cells w23x. The coordinates relative to bregma and the dural surface with the tooth-bar set at zero were as follows: Ž1. A 1.8 mm, L 2.5 mm, V 4.5 and 5.0 mm; Ž2. A 0.6 mm, L 2.0 mm, V 4.0 and 4.5 mm; Ž3. A 0.6 mm, L 3.2 mm, V 4.5 and 5.0 mm w7x. After injection at each site, the syringe was left in place for 3 min before it was retracted. The viability of the cell suspension was maintained at over 70% by the end of the 2 h surgical session. Group CBD received cerebrocortical cell transplants pretreated with BDNF Ž50 ngrml. plus DA Ž10 m M., whereas group CC received only non-treated cerebrocortical cell transplants. The remaining rats received HBSS injections instead of fetal cell suspension Žgroup L, n s 6.. 2.5. TH immunohistochemistry and cell count The procedure for TH immunohistochemistry is fully described elsewhere w24x. Rabbit anti-TH antibody ŽChemicon, USA. and biotinylated goat anti rabbit IgG were used. The peroxidase was visualized by incubation with diaminobenzidine ŽDAB. and hydrogen peroxide. The number of THq neurons was counted in two diametrical 1.82 mm = 16 mm strips, representing 29% of the total surface area of a 16-mm well. The numbers of THq neurons detected were expressed either as the percentages of THq cells found in paired untreated cultures at the end of the treatment period, or as the percentages of the total neuronal cell numbers. The number of THq cells in untreated cultures counted at the end of the experiment was used as the reference value Ži.e. 100%. to determine the percentages in each experiment w23x. The numbers of THq neurons in striatal brain histological sections were also counted. The total number of THq neurons present in the brain striatal region were then estimated according to the formula described previously w1x. This procedure was carried out 12 weeks after fetal nigral cell implantation and 10 weeks after cerebrocortical cell transplantation. 2.6. Statistical analysis Statistical analysis used commercially available statistical software ŽPrimer of Biostatistics 3.01, McGraw Hill Inc... The data were submitted to either a Student t-test or a onertwo-way analysis of variance ŽANOVA. as indicated in Results ŽSection 3.. Either the Dunnett test or the Student-Newman-Keuls test Žas a post-hoc test. was used to compare data samples from the control group with the different treatment groups. Differences were considered significant only when P values were less than 0.05.

46

J. Zhou et al.r DeÕelopmental Brain Research 100 (1997) 43–51

3. Results 3.1. SurÕiÕal of BDNF after addition to primary mesencephalic cultures Since the neurotrophin BDNF is rapidly destroyed by peptidases w26x, it was important to discover how long BDNF persisted in tissue culture medium after its addition as a treatment. Thus, following addition the BDNF protein to the culture medium, its survival was measured by the EIA method in samples taken at different time-points. The results indicated that after BDNF was added, only 20% remained after 1 DIV and these levels were stable until 3 DIV. After the medium changes and the 2nd, 3rd and 4th addition of BDNF were made on the 2nd, 4th and 6th DIV, about 30% of the added BDNF remained detectable in the medium 1 DIV after addition ŽFig. 1.. In untreated mesencephalic cultures, no BDNF was detected in the culture medium at all, indicating that either BDNF was not diffusing out of the tissue, or it was immediately broken down by peptidases. 3.2. Effect of BDNF antibody on mesencephalic neurons in Õitro and in ÕiÕo 3.2.1. Effect of BDNF antibody on the surÕiÕal of dopaminergic neurons in cultures In situ hybridization studies combined with immunocytochemistry for TH have shown the co-existence of BDNF mRNA and TrkB mRNA in developing and adult nigral dopaminergic cells. This has led to the suggestion that BDNF supports the survival of dopaminergic neurons

Fig. 2. Rotation scores of rats receiving a single intranigral microinjection of either turkey anti-BDNF antibody or non-immune IgG. The animals receiving microinjection of anti-BDNF antibody showed ipsilateral rotation after challenge with amphetamine Ž5 mgrkg.. Data are mean"S.E.M. Ž ns 5..

in an autocrine or paracrine manner w11,18x. To test this proposal, specific rabbit anti-BDNF antibody was added to the rat mesencephalic cultures to neutralize BDNF present in the culture. The addition of BDNF antibody reduced the survival of THq neurons by up to 40% during a 7-day observation period when compared with the untreated cultures Žuntreated: 242." 35 neuronsrcm 2 ; BDNF antibody-treated: 143.41 " 13.2 neuronsrcm2 , n s 3.. Addition of non-specific rabbit IgG had no effect on the survival of THq neurons. The antibody had no action on non-dopaminergic neurons present in the same cultures. 3.2.2. Effect of anti-BDNF antibody on the behaÕior of rats receiÕing a single microinjection of BDNF antibody into the substantia nigra Rats receiving a single microinjection of turkey antiBDNF antibody Ž10 m l. into the substantia nigra showed significant ipsilateral rotation 2 weeks after injection ŽTurning: 123 " 26.1 per 30 min., though the number of circling movements declined with time ŽFig. 2.. The rats which received intranigral injections of non-immune rabbit turkey serum showed no significant changes in locomotor behavior. Microinjections of antibody at a location just outside the substantia nigra also caused no significant changes in locomotor behaviour. 3.3. Fetal nigral tissue grafting into 6-OHDA-lesioned rats

Fig. 1. BDNF levels detected by EIA in rat mesencephalic culture medium. Ventral mesencephalon was taken from E14 rat embryos and cultured in serum-free condition for 7 days. BDNF Ž50 ngrml. was added to the mesencephalic cultures on days 1, 3 and 5. Data represent mean"S.E.M. for triplicates from three independent experiments. Some of the standard errors are smaller than the size of individual points and are therefore not visible.

3.3.1. Amphetamine-induced rotation Most rats injected with 6-OHDA, HBSS, or the cultured fetal cells appeared to be healthy and normal during the course of the experiments. In lesioned animals amphetamine-induced ipsilateral turning in 70% of all animals stabilized between 8 and 22 turnsrmin ŽFig. 3.. However, group SNC Žnot BDNFtreated. manifested a rapid decline in amphetamine-induced rotational asymmetry 3 weeks after grafting

J. Zhou et al.r DeÕelopmental Brain Research 100 (1997) 43–51

ŽANOVA: F s 7.37, P - 0.05; Dunnett test: P ) 0.05, compared with the scores before grafting. and showed a further reduction to full compensation Ži.e. - 6 turnsrmin. 6 weeks after transplantation ŽDunnett test: P - 0.05; compared with the scores before grafting; Fig. 3A.. Group SNB ŽBDNF-treated. showed a slightly faster Žasymmetric rotation neutralized in 2.1 weeks. rate of reduction of turning during the 3 weeks after transplantation ŽANOVA: F s 5.21, P - 0.05; Dunnett test: P - 0.05, compared with the scores before grafting; Fig. 3B.. In fact, apart from 2

47

rats which showed higher scores, the other 4 rats in the BDNF-treated group reached full compensation at the 3rd week. Whereas the scores of group L Žbuffer only. showed no significant changes in locomotor behavior over the observation period Žbefore transplantation: 11.1 " 1.5 turnsrmin, mean " S.E.M.; 3 weeks after transplantation: 12.4 " 2.3; 6 weeks: 11.3 " 1.8; 10 weeks: 12.5 " 2.7., thus retaining the full parkinsonian state. 3.3.2. TH immunohistochemistry In group SNC Žfetal nigral neurons, no BDNF treatment., surviving grafts were readily located in the striatum of all animals ŽFig. 4A.. The grafts contained clusters of THq neurons which gave rise to a halo of THq fibers growing into the DA-depleted caudate-putamen, though only part of the denervated host striatum was reinnervated by the grafts. The grafts were estimated to contain between 810 and 1505 Ž1012 " 223, mean " S.E.M.. THq neurons in each rat. These THq neurons were multipolar and of irregular shape ŽFig. 4A.. These surviving THq cells represented between 8% and 15% of the total mesencephalic THq neurons implanted. All fetal SN neurons in group SNB ŽBDNF-treated nigral neurons. were found to survive well as judged by the physical appearance of both isolated neurons and small clusters of THq neurons. The grafted neurons were typically multipolar in shape ŽFig. 4B.. However, pretreatment with BDNF did not produced an increase in the number of THq neurons. The average number of THq neurons in the grafts of group SNB was 211 " 35 which was only about 20.8% of the cell number found in the untreated control group SNC ŽStudent t-test, t s 12.1, P - 0.05., even though at least the same number of dopaminergic neurons were microinjected into each substantia nigra. 3.4. Fetal cortical grafting in 6-OHDA-lesioned rats

Fig. 3. Rotation scores of groups of animals receiving untreated ŽSNC. ŽA. Ž ns 5. or BDNF-treated ŽSNB. ŽB., Ž ns6., before and after transplantation. Data represent mean"S.E.M. Ž`.. Rats, which showed rotation scores of less than 6 turn per min Ždashed line. post-transplantation, were considered to be fully recovered from the behavioral asymmetry. ) indicates significant difference Ž P - 0.05. compared with rotation scores before grafting. Ž‘`’ and ‘e’ represent the scores of each individual animal in the two groups..

3.4.1. Amphetamine-induced rotation The effect of fetal cortical grafts on amphetamine-induced rotation is shown in Fig. 5. Prior to grafting, the three groups ŽL:HBSS only; CC:untreated cortical cells; CBD:BDNFrDA-treated cortical cells. were not different with regard to their mean rotational asymmetries ŽANOVA, P ) 0.5.. Six weeks post-grafting, most animals injected with BDNFrDA-treated cortex Žgroup CBD. displayed a marked, but not statistically significant reduction of amphetamine-induced turning ŽANOVA: F s 3.66, P - 0.05; Dunnett test: P s 0.051 ) 0.05, compared with the scores of pre-transplantation., although the average turning rate was still around 4 turns per min. Turning rates were further reduced to less than 2 turns per min at 10 weeks after transplantation in this group ŽCBD. ŽDunnett test: P 0.05, compared with the scores of pre-transplantation.. In contrast, Group CC did not show any detectable difference from zero time or from the control group ŽL. with HBSS

48

J. Zhou et al.r DeÕelopmental Brain Research 100 (1997) 43–51

injection alone Žtwo-ways of ANOVA, F s 3.25, P ) 0.05.. Group L showed relatively stabilized rotation scores over the four test sessions. 3.4.2. TH immunohistochemistry Microscopic evaluation of host brain sections processed for TH-immunocytochemistry revealed strong vascular responses, and the localized formation of scar tissue around the injection tract. Otherwise the striata in general showed no obvious atrophy or signs of edema. Graft tissue was found in the striatum and around the injection tract and TH neurons were counted in coronal and transverse whole brain sections at 10 weeks, just after asymmetric rotation analysis was completed. Estimates of THq graft neurons numbers at each graft site ranged from 8 to 52 Ž23 " 8, mean " S.E.M., n s 7. in BDNFrDA pretreated cortex Žgroup CBD, Fig. 6B.. Thus between 0.2% to 1.0% of THq neurons implanted survived Žsee Methods, Section 2.. None were detected in recipients of untreated cortical graft Žgroup CC, Fig. 6A.. The presence of THq neurons within the rat striatum can be considered good evidence of survival of the implanted fetal cortical tissue, and suggests that the grafts were the effective agents, and it was not the case, as has been suggested by others, that fetal tissue transplants act merely to stimulate the regeneration and

Fig. 5. Rotation scores of groups L ŽHBSS injection only., CC Žuntreated cortical cells., CBD ŽBDNFrDA-treated cortical cells. recorded at various times before and after transplantation. Data represent mean"S.E.M. Ž ns6–7..

sprouting of host axons, rather than themselves restoring the dopaminergic input to the depleted striatum. Further evidence of this came from the fact that most of these

Fig. 4. TH immunohistochemistry from rat fetal dopaminergic transplants 12 weeks post-grafting. A: THq neurons within grafts of a rat receiving fetal cultured nigral cells without any treatment Žgroup SNC.. B: THq neurons within grafts of a rat receiving fetal cultured nigral cells pre-treated with BDNF Žgroup SNB.. Scale bar: 25 m m.

J. Zhou et al.r DeÕelopmental Brain Research 100 (1997) 43–51

49

Fig. 6. Typical TH immunohistochemistry from single striatal sections of brains containing rat fetal cerebrocortical transplants without treatment ŽA., or pretreated with BDNFrDA ŽB.. Details of the preparation were described in the text. Scale bar: A, 50 m m; B, 250 m m. cc: corpus callosum; G: graft.

control animals Ž5 out of 6 in group CC. did not display any sign of decreased rotational behavior.

4. Discussion The neurotrophin BDNF has been shown in this and other laboratories w11,24x to exert trophic effects on dopaminergic fetal neurons. The results of the present study suggest that BDNF may be essential for the survival of both fetal and adult dopaminergic neurons. Thus, addition of antibodies to BDNF led to the death of 40% of fetal dopaminergic neurons in culture and death of nigral dopaminergic neurons when microinjected in vivo as evidenced by the induction of asymmetric locomotion. BDNF, therefore, not only promotes the survival of nigral dopaminergic neurons in vitro, it also co-induces Žwith dopamine. expression of the full dopaminergic phenotype in fetal cerebral cortex when added to cultures of these neurons, together with dopamine. Published studies have shown that in the absence of added peptidase inhibitors, BDNF is rapidly destroyed when being extracted from brain tissue w14,26x. It was therefore important to establish whether BDNF added to cultures was also rapidly destroyed in the absence of peptidase inhibitors, having, therefore, only brief contact with the cells, perhaps for only minutes or seconds. The data reported here show that only 20–30% of added BDNF survives when added to cultures at 50 ngrml. Thus 70–

80% is destroyed by peptidase action, possibly within minutes. The maximum survival time being 1 DIV in the present studies, greater time-resolution studies being necessary to establish the precise rate of degradation of added BDNF. The findings reported here show that persisting ambient concentrations of only between 10 ; 20 ngrml BDNFrml culture medium, are enough to produce the profound actions of BDNF added to primary cultures i.e. prolongation of dopaminergic neurons survival times and dopaminergic phenotype induction. Others have reported that 10 ng BDNFrml was enough to cause more than 250% increase in survival of THq cells in mesencephalic culture w20x. 4.1. Actions of BDNF antibody in Õitro and in ÕiÕo Addition of a highly specific BDNF antibody Žsee Methods. to fetal mesencephalic dopaminergic cultures caused 40% loss of THq neurons over 7 days. This was not due to non-specific toxicity, as the antibody had no action on non-dopaminergic neurons present in the cultures, and non-immune rabbit IgG had no action on any neurons present. Indeed the presence of BDNF mRNA has been shown in the embryonic rat brain and postnatal and adult substantia nigra w8,9,12x. We have previously reported an intracellular level of BDNF of immediately postnatal rat brain of 2.96 " 0.31 ngrg tissue and adult rat brain of 4.38 " 0.31 ngrg tissue in ventral midbrains w26x. Thus, it seems probable that neutralizing this endogenous

50

J. Zhou et al.r DeÕelopmental Brain Research 100 (1997) 43–51

BDNF with added specific BDNF antibody leads to death of dopaminergic neurons, indicating the essential requirement of BDNF for prolongation of the life of these neurons, but of not other neurons present in the cultures. Since unilateral intra-nigral microinjections of this same antibody induced ipsilateral rotations, similar to those caused by 6-OHDA, we conclude that specific destruction of nigral dopaminergic neurons was caused by the highly specific BDNF antibody. This indicates the obligatory presence of BDNF required for in vivo survival of these neurons. Since non-immune serum produced no effects, non-specific toxicity of the serum was clearly not responsible for the observed effects. Microinjection of the specific BDNF serum outside the substantia nigra also produced no effects, showing that the surgical procedures and microinjections themselves were not the cause of the actions of the BDNF antiserum. These results suggest that early loss of BDNF should be considered as a possible causative factor in the events leading to the onset of Parkinson’s disease.

trol untreated cerebral cortex is ineffective Žsee also w22x.. Immunohistochemistry showed the presence of THq neurons in sections containing grafts of BDNFrDA-treated cortical tissue 10 weeks after grafting, as expected, no THq neurons were seen in sections of untreated cortical graft tissue. A maximum of about 1.04% and an average of 0.5% of the total number of THq neurons implanted into the striatum had survived at 10 weeks, i.e. an estimated maximum of 52 out of 5 = 10 3 of THq neurons at each graft site. This survival of dopaminergic neurons, artificially induced in cerebral cortex, shows the robust and stable nature of this induced phenotype. It suggests that human fetal cerebral cortex containing induced dopaminergic neurons could provide a large source of graft tissues for surgical implantation to treat Parkinson’s disease. Such an approach using tissue culture combined with cryostorage would overcome the major problem of mesencephalic fetal tissue supply currently facing this form of therapy of Parkinson’s disease w2,23,25x.

4.2. Nigral grafts in 6-OHDA-lesioned rats Grafts of BDNF-treated cultured mesencephalic dopaminergic tissue placed in the striatum of 6-OHDA-lesioned rats caused recovery from asymmetric rotation induced by amphetamine by 3 weeks. This was significantly faster than for animals receiving grafts of untreated cultured fetal nigral cells, i.e. recovery in 2.1 weeks earlier due to treated grafts. Thus there appeared at first to be some benefit in treating the graft tissue with BDNF before implantation Ži.e. 20% faster.. However, immunohistochemistry showed 80% fewer BDNF-treated dopaminergic neurons were surviving 8 weeks after grafting compared with untreated tissues. Thus, the true functional effectiveness of BDNF-treated THq mesencephalic neurons could be estimated as 5-fold greater than untreated neurons rather than just 20% as effective. Why there is a lower survival rate of BDNF-treated THq nigral cells is not clear. Perhaps they do not survive harvesting so well, or are more vulnerable to death in the hypoxic in vivo striatal environment when present as a graft of harvested cells. This contrasts with the 200% promotion of survival of these same BDNF-treated dopaminergic neurons in tissue culture w24x.

4.4. SurÕiÕal of implanted TH q neurons and functional recoÕery One earlier report estimated that more than 200 fetal ventral mesencephalic THq neurons ventral mesencephalic implanted into the rat striatum are required for full compensation of asymmetric rotation w17x. In the present study we introduced about 3 = 10 4 THq cortical neurons and 10 4 ventral mesencephalic THq neurons which had been cultured and harvested as cell suspensions, of which 0.2 to 1.0% of the former, and 8–15% of the latter survived. This very rough estimate of total THq neurons surviving is based on the method of Ambercrombie et al. w1x, developed to estimate populations of nuclei, and could well underestimate THq cell numbers by ten-fold w1x. If the estimate is fairly accurate, then dopaminergic neurons induced in cerebral cortex are clearly very effective in neutralizing asymmetric rotation ŽFig. 5.. This laboratory has previously reported the more robust survival of cultured fetal cerebral cortex, over ventral mesencephalic tissue when used as grafts delivered as cell suspensions into rat striatum w22x.

4.3. Cortical grafts in 6-OHDA-lesioned rats

Acknowledgements

This laboratory has previously published the finding that primary cultures of both human and rat fetal cerebral cortex harvested and grafted into the rat striatum provide robust, long-lasting grafts which do not reduce or prevent asymmetric rotation in 6-OHDA-lesioned rats w22x. We now reported that fetal cerebral cortex co-treated with BDNF and dopamine to induce the full dopaminergic phenotype w23,25x, can prevent this amphetamine-induced asymmetric rotation in 6-OHDA-treated rats, whereas con-

We thank Dr. J.A. Miller of Amgen, Thousand Oaks, CA, USA for kindly providing BDNF antibodies and Dr. R.M. Lindsay of Regeneron, Pharmaceuticals, Tarrytown, New York, USA for generously providing human recombinant neurotrophins. We thank Dr. S.B. Dunnett of Cambridge University, UK and Dr. D. Clark of Oxford University, UK for help in setting up the computerized and automated rotometer apparatus. We thank Mr. Glyn Millhouse of this Department for excellent photographic assis-

J. Zhou et al.r DeÕelopmental Brain Research 100 (1997) 43–51

tance. This work was supported by grants from the Parkinson’s Disease Society of the United Kingdom.

References w1x Abercrombie, M., Estimation of the nuclear population from microtome sections, Anat. Rec., 94 Ž1946. 239–247. w2x Ahlskog, J.E., Cerebral transplantation for Parkinson’s disease: current progress and future prospects, Mayo Clin. Proc., 68 Ž1993. 578–591. w3x Bjorklund, A., Better cells for brain repair, Nature, 362 Ž1993. ¨ 414–415. w4x Du, X.Y. and Iacovitti, L., Synergy between growth factors and transmitters required for catecholamine differentiation in brain neurons, J. Neurosci., 15 Ž1995. 5420–5427. w5x Du, X.Y., Stull, N.D. and Iacovitti, L., Brain-derived neurotrophic factor work coordinately with partner molecules to initiate tyrosinehydroxylase expression in striatal neurons, Brain Res., 680 Ž1995. 229–233. w6x Dunnett, S.B. and Bjorklund, B. Ž1991.. Neural Transplantation: Practical Approach. IRL. London. w7x Forni, C., Brundin, P., Strecker, R.E., Elganouni, S., Bjorklund, A. ¨ and Nieoullon, A., Time-course of recovery of dopamine neuron activity during reinnervation of the denervated striatum by fetal mesencephalic grafts as assessed by in vivo voltammetry, Exp. Brain Res., 76 Ž1989. 75–87. w8x Friedman, W.J., Olson, L. and Persson, H., Cells that express brain-derived neurotrophic factor mRNA in the developing postnatal rat brain, Eur. J. Neurosci., 3 Ž1991. 688–697. w9x Gall, C.M., Gold, S.J., Isackson, P.J. and Seroogy, K.B., Brain-derived neurotrophic factor and neurotrophin 3 mRNAs are expressed in ventral midbrain regions containing dopaminergic neurons, Mol. Cell. Neurosci., 3 Ž1992. 56–63. w10x Hyman, C., Hofer, M., Barde, Y.-A., Juhasz, M., Yancopoulos, G.D., Squinto, S.P. and Lindsay, R.M., BDNF is a neurotrophic factor for dopaminergic neurons of the substantia nigra, Nature, 350 Ž1991. 230–232. w11x Hyman, C., Juhasz, M., Jackson, C., Wright, P., Ip, N.Y. and Lindsay, R.M., Overlapping and distinct action of the neurotrophins BDNF, NT-3, and NT-4r5 on cultured dopaminergic and GABAergic neurons of the ventral mesencephalon, J. Neurosci., 14 Ž1994. 335–347. w12x Maisonpierre, P.C., Belluscio, L., Friedman, B., Alderson, R.F., Wiegand, S.J., Furth, R.F., Lindsay, R.M. and Yancopoulos, G.D., NT-3, BDNF, and NGF in the developing rat nervous system: parallel as well as reciprocal patterns of expression, Neuron, 5 Ž1990. 501–509. w13x Mayer, E., Fawcett, J.W. and Dunnett, S.B., Basic fibroblast growth factor promotes the survival of embryonic ventral mesencephalic

w14x

w15x

w16x

w17x

w18x

w19x

w20x

w21x

w22x

w23x

w24x

w25x

w26x

51

dopaminergic neurons-II. effects on nigral transplants in vivo, Neuroscience, 56 Ž1993. 389–398. Nawa, H., Carnahan, J. and Gall, C., BDNF protein measured by a novel enzyme immunoassay in normal brain and after seizure: partial disagreement with mRNA levels, Eur. J. Neurosci., 7 Ž1995. 1527–35. Rosenfeld, R.D., Zeni, L., Haniu, N., Talvenheimo, J., Radka, S.F., Bennett, L., Miller, J.A. and Welcher, A.A., Purification and identification of brain-derived neurotrophic factor from human serum, Protein Express. Purific., 6 Ž1995. 465–471. Sanchez-Ramos, J.R., Michel, P., Weiner, W.J. and Hefti, F., Selective destruction of cultured dopaminergic neurons from fetal rat mesencephalon by 1-methyl-4-phenylpyridinium:Cytochemical and morphological evidence, J. Neurochem., 50 Ž1988. 1934–44. Schmidt, R.H., Bjorklund, A., Stenevi, U., Dunnett, S.B. and Gage, ¨ F.H., Activity of intrastriatal nigral suspension implants as assessed by measurements of dopamine synthesis and metabolism, Acta Physiol. Scand. (Suppl.), 522 Ž1983. 19–28. Seroogy, K.B., Lundgren, K.H., Tran, T.M., Guthrie, K.M., Isackson, P.J. and Gall, C.M., Dopaminergic neurons in rat ventral midbrain express brain-derived neurotrophic factor and neurotrophin-3 mRNAs, J. Comp. Neurol., 342 Ž1994. 321–34. Shults, C.W., Matthews, R.T., Altar, C.A., Hill, L.R. and Langlais, P.J., A single intramesencephalic injection of brain-derived neurotrophic factor induces persistent rotational asymmetry in rats, Exp. Neurol., 125 Ž1994. 183–94. Spenger, C., Hyman, C., Studer, L., Egli, M., Evtouchenko, L., Jackson, C., Dahljorgensen, A., Lindsay, R.M. and Seiler, R.W., Effects of BDNF on dopaminergic, serotonergic, and GABAergic neurons in cultures of human fetal ventral mesencephalon, Exp. Neurol., 133 Ž1995. 50–63. Ungerstedt, U. and Arbuthnott, G.W., Quantitative recording of rotational behaviour in rats after 6-hydroxydopamine lesions of the nigrostriatal dopamine system, Brain Res., 24 Ž1970. 485–493. Walters, A.M., Clarke, D.J., Bradford, H.F. and Stern, G.M., The properties of cultured fetal human and rat brain tissue and its use as grafts for the relief of the parkinsonian syndrome, Neurochem. Res., 17 Ž1992. 893–900. Zhou, J., Bradford, H.F. and Stern, G.M., Induction of dopaminergic neurotransmitter phenotype in rat embryonic cerebrocortex by the synergistic action of neurotrophins and dopamine, Eur. J. Neurosci., 8 Ž1996. 2328–2339. Zhou, J., Bradford, H.F. and Stern, G.M., The response of human and rat fetal ventral mesencephalon in culture to the brain-derived neurotrophic factor treatment, Brain Res., 656 Ž1994. 147–156. Zhou, J., Bradford, H.F. and Stern, G.M., The stimulatory effect of brain-derived neurotrophic factor on dopaminergic phenotype expression of embryonic rat cortical neurons in vitro, DeÕ. Brain Res., 81 Ž1994. 318–324. Zhou, J., Pliego-Rivero, B., Bradford, H.F. and Stern, G.M., The BDNF content of postnatal and adult rat brain: the effects of 6-hydroxydopamine lesions, DeÕ. Brain Res., 97 Ž1996. 297–303.