©2005 FASEB The FASEB Journal express article 10.1096/fj.05-3845fje. Published online July 21, 2005. Haploinsufficiency for trkB and trkC receptors induces cell loss and accumulation of α-synuclein in the substantia nigra Oliver von Bohlen und Halbach,* Liliana Minichiello,† and Klaus Unsicker* *Interdisciplinary Center for Neurosciences, Department of Neuroanatomy, University of Heidelberg, 69120 Heidelberg, Germany; and †European Molecular Biology Laboratory, 00016 Monterotondo, Italy Corresponding author: Oliver von Bohlen und Halbach, Interdisciplinary Center for Neurosciences (IZN), Department of Neuroanatomy, University of Heidelberg, Im Neuenheimer Feld 307, D-69120 Heidelberg, Germany. E-mail: O.von_Bohlen@web.de ABSTRACT The neurotrophins brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT-3) have been shown to promote survival and differentiation of midbrain dopaminergic (DAergic) neurons in vitro and in vivo. This is consistent with their expression and that of their cognate receptors, trkB and trkC, in the nigrostriatal system. Degeneration of DAergic neurons of the substantia nigra and α-synuclein-positive aggregates in the remaining substantia nigra (SN) neurons are hallmarks of Parkinson’s disease (PD). Reduced expression of BDNF has been reported in the SN from PD patients. Moreover, mutations in the BDNF gene have been found to play a role in the development of familial PD. We show now that haploinsufficiencies of the neurotrophin receptors trkB and/or trkC cause a reduction in numbers of SN neurons in aged (21-23 month old) mice, which is accompanied by a reduced density in striatal tyrosine hydroxylase immunoreactive (TH-ir) fibers. These aged mutant mice, in contrast to wild-type littermates, display an accumulation of α-synuclein in the remaining TH-positive neurons of the SN. We conclude that impairment in trkB and/or trkC signaling induces a phenotype in the aged SN, which includes two hallmarks of PD, losses of TH positive neurons and axons along with massive neuronal deposits of α-synuclein. Key words: dopamine ● mice ● neurotrophins ● striatum T he neurotrophins brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT-3), along with their cognate receptors trkB and trkC, respectively, are expressed in the substantia nigra pars compacta (SNpc) and in the striatum, both during development and adulthood (1–3). In humans, a majority of dopaminergic (DAergic) neurons of the SNpc shows immunoreactivities for trkB (71%) or trkC (86%; ref 3). Both BDNF and NT-3 also promote survival and differentiation of DAergic neurons in vivo (4), lending support to speculations that impaired signaling through their cognate receptors may be harmful for DAergic neurons. Lesions of the DAergic nigrostriatal system with 6-hydroxydopamine (6-OHDA) have been shown to reduce BDNF mRNA levels in the SN of adult rats (5). Intrastriatal grafts of fibroblasts genetically engineered to produce BDNF partially prevent the loss of nerve terminals and Page 1 of 19 (page number not for citation purposes) completely prevent the loss of cell bodies of the 6-OHDA-lesioned nigrostriatal system (6). Further along this line, 6-OHDA-induced rotational behavior can be prevented by BDNF somatic gene transfer into neurons of the SN (7). In the MPTP model of Parkinson’s disease (PD), implantation of immortalized fibroblasts overexpressing BDNF close to the SN largely prevents MPTP-induced DAergic neuronal degeneration and enhances DA levels (8). Together, these data suggest that neurotrophins are important factors for the maintenance and survival of DAergic neurons and that dysfunctions in neurotrophin signaling may cause pathological alterations in the DAergic nigrostriatal system. Postmortem analyses of PD-diseased human SN have revealed a reduction in BDNF mRNA and protein (9–11), raising the possibility that there may be a link between reduced levels of BDNF and PD. In addition, a recent publication (12) has suggested that pathogenic α-synuclein mutations (A30P or A53T) may be linked to a loss in BDNF production. We report now that aged mice that are heterozygous for the trkB or/and trkC receptor genes show losses of DAergic neurons and their axons projecting to the striatum. Virtually all remaining DAergic cell bodies display accumulations of α-synuclein. Thus, partial ablation of trkB or/and trkC genes generates a phenotype, which includes several hallmarks of PD, suggesting that endogenous ligands to these receptors must be crucial for the maintenance and metabolic welfare of the nigrostriatal DAergic system during aging. MATERIALS AND METHODS Animals and tissue processing Male heterozygous aged (21-23 months old) trkB(+/−), trkC(+/−), trkB/trkC(+/−)/(+/−) mice, and agematched littermates (trkB/trkC(+/+)/(+/+)), as well as adult (6-8 months old) heterozygous trkB(+/−), trkC(+/−), trkB/trkC(+/−)/(+/−) mice, and control littermates (trkB/trkC(+/+)/(+/+)) were used and maintained in accordance with the institutional guidelines for animal welfare. By crossbreeding heterozygous trkB and heterozygous trkC mice (13) and, thereafter, by breeding heterozygous trkB/C mice, trkB(+/−), trkC(+/−), trkB/C(+/−)/(+/−) mutant mice, and control littermates (trkB/C(+/+)/(+/+)) were obtained. The genotypes were determined by PCR from tail biopsies of the offsprings. The trkB genotype was determined by PCR amplification using a common 5′ primer (5′-TCG CGT AAA GAC GGA ACA TGA TCC-3′) and either a 3′ primer for the wildtype allele (5′-AGA CCA TGA TGA GTG GGT CGC C-3′) or a 3′ primer from the pgk-1 promotor of the neo cassette (5′-GAT GTG GAA TGT GTG CGA GGC C-3′). The trkC genotype was determined using a common 5′ primer (5′-CTG AAG TCA CTG GCT AGA GTC TGG G-3′) and either a 3′ primer for the wild-type allele (5′- GTC CCA TCT TGC TTA CCC TGA GG-3′) or a 3′ primer from the pkg-1 promotor of the neo cassette (5′-CCA GCC TCT GAG CCC AGA AAG C-3′). The PCR amplified DNA was analyzed on a 1.5% agarose gel. Animals were deeply anesthetized with ether and transcardially perfused with phosphate buffered saline (PBS: 2.0 g NaH2PO4, 10.73 g Na2HPO4 and 9.0 g NaCl in 1000 ml Aqua dest; pH 7.2) and afterward with fixative (PBS containing 4% formaldehyde, derived from paraformaldehyde). The brains were left in situ for half an hour at room temperature and then removed. Brains were immersed in the same fixative for 4-7 days. Serial coronal sections were made using a vibratome (VT1000E, Leica, Germany) and mounted on gelatin-coated slides. Page 2 of 19 (page number not for citation purposes) Staining protocols Nissl-staining Mounted sections were rinsed 5 min in distilled water and afterward transferred to a 1.5% cresylviolet solution. After this, sections were incubated in a solution containing 2.5 ml glacial acetic acid, 40 ml ethanol, and 207.5 ml distilled water. Sections were then transferred to an ascending alcohol-series (70, 90, 96% ethanol), and finally to xylol, before they were coverslipped using Merckoglas (Merck, Germany). Immunohistochemistry Serial sections were alternately stained with TH antibodies, α-synuclein antibodies, or doublelabeled for both TH and α-synuclein. For single-labeling with polyclonal rabbit TH antiserum, sections were rinsed, preincubated in a solution containing 3% normal goat serum (NGS) and 0.3% Triton-X in PBS. Thereafter, sections were incubated in a solution containing 1% NGS, 0.3% Triton-X in PBS, and polyclonal rabbit TH antiserum (Chemicon; 1:200) for 48 h. Sections were washed three times in PBS and then transferred to a solution containing biotinylated anti-rabbit IgG (Vector; 1:200), washed three times in PBS, and subsequently incubated in a solution containing Cy3-conjugated streptavidin (Jackson ImmunoResearch: 1:1500). Sections were washed in PBS, counter-stained with a solution containing DAPI (Molecular Probes; 1:15000), washed again, and coverslipped in fluorescent mounting medium (DAKO). Alpha-synuclein was visualized using the following protocol: mounted sections were preincubated in a solution containing 3% NGS and 0.3% Triton-X in PBS. Thereafter, sections were incubated in a solution containing 1% NGS, 0.3% Triton-X in PBS and polyclonal rabbit αsynuclein antiserum (Chemicon; 1:1500). Sections were washed two times in PBS and then transferred to a solution containing biotinylated anti-rabbit IgG (Vector; 1:200), washed two times in PBS, and subsequently incubated in a solution containing Cy3-conjugated streptavidin (Jackson ImmunoResearch; 1:1000). Sections were washed in PBS, counter-stained with a solution containing DAPI (Molecular Probes; 1:15000), washed again, and coverslipped in fluorescent mounting medium (DAKO). For double-labeling, sections were preincubated in a solution containing 3% NGS and 0.3% Triton-X in PBS. Thereafter, sections were incubated in a solution containing 1% NGS, 0.3% Triton-X in PBS and polyclonal sheep TH antiserum (Chemicon) for 48 h. Sections were washed three times in PBS and then transferred to a solution containing anti-sheep Alexa Fluor 488 (Molecular Probes; 1:200), washed three times in PBS, and then incubated in polyclonal rabbit α-synuclein antiserum (1:1500). Anti-α-synuclein binding was visualized using biotinylated secondary anti-rabbit IgGs and Cy3-streptavidin, according to the procedure described above. Sections were washed in PBS, counter-stained with a solution containing DAPI (Molecular Probes; 1:15,000), washed, and coverslipped in fluorescent mounting medium (DAKO). Page 3 of 19 (page number not for citation purposes) Thioflavin S staining For thioflavin S staining, fixed sections of 21-month-old heterozygous trkB/trkC(+/−)/(+/−) mice (n=2) and age-matched littermates (trkB/trkC(+/+)/(+/+); n=2) were incubated with 0.05% thioflavin S (Sigma) for 8 min and washed three times with 80% ethanol for 5 min (14). Every second section was immunostained against TH, as outlined above, counter-stained with DAPI, washed, and coverslipped in fluorescent mounting medium (DAKO). Quantification All measurements were done in a blind study. For shrinkage correction of the subsequent measurements, the thickness of the original section thickness (after cutting) and the thickness of the section after staining and embedding were measured using the commercially available stereological system StereoInvestigator (MicroBrightField) and a computer-driven motorized stage (Merzhäuser). The subsequent measurements were corrected by these factors. In the case of Nissl staining, the instrumental thickness was 30 µm and the postprocessing thickness was 22.7 ± 0.4 µm, and in the case of the immunohistochemical staining, the instrumental thickness was 25 µm and the postprocessing thickness was 19.6 ± 0.7 µm. Optical fractionator To estimate total neuronal numbers, an unbiased counting system, the optical fractionator (15) was used. Total numbers of SNpc neurons on Nissl-stained sections were estimated using StereoInvestigator in one brain hemisphere. A systematic random series of sections throughout the SNpc was collected. The first sampled section was selected randomly by using the random number generator of StereoInvestigator. On each section, the outlines of the borders of the SNpc were drawn with ×5 lens and a grid was placed randomly over the area of interest. At predetermined positions of the grid, cells were counted within three-dimensional optical disectors. The total number of neurons was estimated using the equation: N = Q- × t/h × 1/asf × 1/ssf where asf is the area sampling fraction, h is the height of the sampling fraction, N is the total neuronal number, Q- is the number of objects counted, ssf is the section sampling fraction, and t is the section height. Sampling was done using an oil objective with a magnification of ×63 and a numerical aperture of 1.25. In all cases the following stereological parameters were used: size of guard zones = 2 µm, counting frame area = 2025 µm2, sampling grid area = 10,000 µm2. The biological variance (BV) or inter-animal variance as well as the coefficient of error (CE, a value that is indicative of the intra-animal variance) was calculated. As CE values do not exceed BV values, precision of the estimates is considered to be sufficient (15). Abercrombie’s formula Total numbers of TH-positive neurons within the SNpc in one brain hemisphere were estimated using the Abercrombie Eq. (16). This method renders biases within the range of the optical dissector by taking into account that the particles counted are small compared with the section Page 4 of 19 (page number not for citation purposes) thickness (17). Cell counts of TH-immunoreactive (TH-ir) perikarya within the SNpc were made from every sixth section, using a ×40 objective and the image processing and analysis program ImageTool 3.0 (University of Texas Health Science Center, San Antonio, TX). To account for the overcount due to the presence of split particles in the sections, counts were corrected using the Abercrombie’s correction formula. The Linderstrom-Lang/Abercrombie (LLA) equation for estimating numerical neuronal densities is: N = n*t(t+H) or N/n = f = t/(t+H) N is an estimate of the number of objects in the defined region, n is the counted number of objects, t is the mean thickness of the virtual section, H is the mean height of the objects, and f is the conversion factor for converting n to N. In a first step, n was counted. In a second step, H, the height of the cells in the z-axis, was estimated using computer-driven motorized stage (Merzhäuser, Germany) under the control of StereoInvestigator. As in case of the optical fractionator, the mean, the standard deviation (SD), CE, and BV were calculated. Optical density Densitometric analysis of the TH staining in the striatum was performed with the aid of Image J (NIH). For the evaluation of optical densities (OD), aged (21 months old) heterozygous trkB/trkC(+/−)/(+/−) mice (n=3), age-matched littermates (trkB/trkC(+/+)/(+/+); n=4), and adult control mice (trkB/trkC(+/+)/(+/+); n=3) were used. Six consecutive sections were analyzed per animal. The staining intensities were displayed as arbitrary OD values (18). The obtained values were averaged and expressed as a percentage of control values. Relative fiber densities The relative fiber densities of TH-positive fibers in the striatum were quantified as described previously (19). Immunostained fibers were visualized using a ×63 oil objective (numerical aperture 1.25). The region of interest (ROI) of one focal plane was captured by an Axiocam digital camera (Zeiss, Germany) mounted on a microscope (Axioplan 2 Imaging) under the control of the software Axiovision (Version 3.1). A grid consisting of single pixels spaced 2.5 × 2.5 µm apart (in the x- by y-plane) was overlaid on the computer-stored image. Fibers intercepting the grid-points were calculated. Relative fiber densities (Q) were given as quotient of grid points, intercepted by fibers (Gi) divided by the total numbers of grid points (Go). At least six different ROIs at different rostro-caudal positions of the striatum were analyzed, spaced 120 µm apart. Data were obtained from adult (6-8 months old) heterozygous trkB(+/−) (n=4), trkC(+/−) (n=3), trkB/trkC(+/−)/(+/−) mice (n=3), and control littermates (trkB/trkC(+/+)/(+/+); n=4) and from aged (21-23 months old) trkB(+/−) (n=3), trkC(+/−) (n=3), trkB/trkC(+/−)/(+/−) mice (n=3), and age-matched littermates (trkB/trkC(+/+)/(+/+); n=3). Estimates of area-densities of α-synuclein containing neurons For this analysis, a microscope (Axioplan 2 Imaging) with a computer-driven digital-camera (Axiocam) and a ×40 objective was used. Page 5 of 19 (page number not for citation purposes) To count α-synuclein-accumulating neurons, areas of interest (representing a window of 10,000 µm2) were delineated within the SNpc, which was identified by the help of the TH immunostaining. Digital image analysis-assisted counts of profiles were made using ImageJ 2.19 (NIH). According to their appearance, the α-synuclein immunoreactive cells were classified into two groups (20): type 1 displays diffuse cytoplasmatic staining or irregularly shaped staining of moderate intensity. Type 2 is characterized by a discrete staining corresponding to “pale bodies” and to typical Lewy bodies. Data were obtained from adult (6-8 months old) heterozygous trkB(+/−) (n=4), trkC(+/−) (n=3), trkB/trkC(+/−)/(+/−) mice (n=3), and control littermates (trkB/trkC(+/+)/(+/+);n=3) and from aged (2123 months old) trkB(+/−) (n=3), trkC(+/−) (n=3), trkB/trkC(+/−)/(+/−) mice (n=3), and age-matched littermates (trkB/trkC(+/+)/(+/+); n=3). ANOVA with Tukey-Kramer post-hoc test was used for statistical evaluation. P values <0.05 were considered as statistically significant. RESULTS Neuronal numbers within the SNpc Since aging has been reported to be accompanied by a numerical reduction of neurons within the human SNpc (21), we first compared neuronal numbers in the SNpc of adult and aged mice (6-8 and 21-23 months of age, respectively). Figure 1A reveals that aged mice showed a significant 13.5% reduction in total neuronal numbers. We next analyzed total neuronal numbers in the SNpc of 21-23 months old trkB(+/−), trkC(+/−), and trkB/trkC(+/−)/(+/−) mice and respective littermates (Fig. 1A; Table 1). Compared with age-matched trkB/C(+/+)/(+/+) mice, haploinsufficiency for trkB resulted in a statistically significant decrease of total neuron numbers by 15% in the aged animals. A further significant decrease was found in trkB/trkC(+/−)/(+/−) mice (21.3%, P<0.01, ANOVA). Numbers of dopaminergic neurons within the SNpc We next analyzed numbers of TH-immunopositive SNpc neurons in adult and aged wild-type mice. Table 2 reveals a decrease in total neuronal numbers in the SNpc of aged mice, compared with adult mice. This was paralleled by a statistically significant decrease of 16.6% in the numbers of DAergic neurons visualized by TH-ir. Comparisons of TH-ir neuron numbers in the SNpc revealed slight but insignificant differences were found between adult wild-type and adult mutant littermates (Fig. 1B; Table 2). Comparisons of TH-ir neuron numbers in the SNpc of aged heterozygous trk mutants and respective control mice revealed significant reductions ranging from -14.4% (aged trkC(+/−)) and 17.8% (aged trkB(+/−)) to -26.0% (aged trkB/trkC(+/−)/(+/−)) (Fig. 1C; Table 2). Compared with adult trkB(+/−) mice, aged trkB(+/−) mice displayed a reduction of 27.8%, while aged trkC(+/−) mice displayed a 25.7% reduction as compared with adult trkC(+/−) mice. The most prominent reduction was found in double-heterozygous animals. Compared with adult Page 6 of 19 (page number not for citation purposes) trkB/C(+/−)/(+/−) mice, aged trkB/C(+/−)/(+/−) mice showed a reduction of 34.5%. Together, these data indicate that both trkB and trkC haploinsufficiencies cause prominent reductions in numbers of total and TH-positive neurons in the SNpc of aged but not adult mice. Relative densities of TH-ir striatal fibers Having shown that impaired trk signaling reduces the numbers of TH-positive neuronal cell bodies in the SNpc, we next analyzed the striatum. Analysis of the staining intensities revealed a significant 18% age-related reduction in staining intensities in trkB/trkC(+/+)/(+/+) mice. In aged animals, there was a reduction in optical densities of ~19% by comparing aged trkB/trkC(+/−)/(+/−) with their age-matched control littermates. Reduction in TH immunostaining in the striatum can reflect either a reduction in the number of TH-positive fibers or a down-regulation of TH in dopaminergic fibers (18). Therefore, we studied possible alterations in the densities of striatal TH-ir fibers in adult (6-8 months old) and aged (21-23 months old) wild-type and mutant mice. There is a significant age-related (6-8 vs. 21-23 months) decline in the density of TH-positive fibers in normal (trkB/trkC(+/+)/(+/+)) mice of 18.8%. In adult mutant animals, as compared with age-matched wild-type littermates, slight but insignificant, reductions in the densities of TH-ir fibers were noted in trkB(+/−) and trkC(+/−) mice. A significant reduction (-8.8%) was only found in the heterozygous double mutant mice (Fig. 1D). Significant reductions in the densities of TH-ir fibers were found in aged trkB or/and trkC heterozygous knockout mice, compared with age-matched (21-23 months old) wild-type littermates (Fig. 1E). The most pronounced reduction in TH-ir fiber densities was found in aged trkB/trkC(+/−)/(+/−) mice (-26% compared with age-matched controls). Together, these data indicate that losses of TH-ir fibers in the striatum of trkB or/and trkC heterozygous mutant mice parallel losses of TH-positive neuronal cell bodies in the SNpc. Abnormal accumulation of α-synuclein in SNpc neurons In addition to the prominent degeneration of DAergic neurons in the SNpc of PD patients, abnormal accumulations of α-synuclein in neuronal cell bodies of the SNpc are morphological hallmarks of PD (22). Figure 2A and B summarizes and Fig. 3A–C shows immunoreactivity for α-synuclein accumulating in the neuronal cell bodies of the SNpc of trkB or/and trkC mutants mice, similar to what has been described for α-synucleinopathies (20, 23). Adult and aged control mice show no or very rare deposits of α-synuclein in the SNpc. Furthermore, adult (6-8 months old) heterozygous trkB, trkC, and trkB/C mice also display only rare deposits of αsynuclein in the SNpc (Fig. 2A). A marked and significant increase in profiles with cytoplasmatic α-synuclein accumulation in the SNpc was found in the aged heterozygous mutant mice; the highest increase was found in the aged trkB/trkC(+/−)/(+/−) mice (Fig. 2B). Double-labeling experiments, using antibodies directed against TH and α-synuclein (Fig. 3D–L), revealed that in the aged double heterozygous knockout mice 95.2% of the α-synuclein accumulating cell bodies in the SNpc were TH positive, whereas only 4.8% of the α-synuclein accumulating cells were Page 7 of 19 (page number not for citation purposes) TH negative. In aged heterozygous trkB and trkC mice 4 and 3%, respectively, of the αsynuclein accumulating cells were TH negative. These TH-negative α-synuclein-accumulating cells may represent non-DAergic cells or DAergic neurons, which have lost their DAergic phenotype. Thus, losses of TH-positive neurons in mice lacking one allele of trkB or/and trkC are accompanied by massive accumulations of α-synuclein in remaining TH-ir cell bodies in the SNpc. Alpha-synuclein-ir structures in the human SN have been classified into two morphological types (20). With the use of this classification, only type 1 but no type 2 α-synuclein accumulating neurons were found in trkB or/and trkC mutant mice. In addition, neither aged heterozygous trkB/trkC(+/−)/(+/−) mice (Fig. 4) nor age-matched littermates (trkB/trkC(+/+)/(+/+)) displayed thioflavin S positive inclusions in the SNpc. DISCUSSION Effect of aging on the DAergic system It has been firmly established that numbers of DAergic neurons in the human SN decrease with age, with an approximate 5-10% loss per decade (21). Significant age-related losses of TH-ir neurons in the SN have also been reported for monkeys (24) and mice (25, 26), which are accompanied by a reduction of TH-ir fiber densities in the hippocampus and amygdala (19). Thus, our present data corroborate the notion that significant losses of nigral neurons occur during normal aging. In addition, we could demonstrate that this loss of neurons is accompanied by a decline in TH-ir fiber densities in the striatum but not by an accumulation of α-synuclein in the somata of nigral neurons in the aged wild-type mice. Reduced neurotrophin-signaling contributes to generate a PD-like phenotype In humans, losses of nigral neurons are dramatically enhanced in PD. In addition, there is evidence to suggest that striatal DA-levels decline with age (27, 28) and that these alterations are aggravated during the progressive degeneration of the DAergic system in PD (23). Degeneration of the DAergic neurons of the SNpc and a subsequent loss of DAergic nerve terminals in the striatum are responsible for motor disturbances in PD; these symptoms, however, become only apparent when ~70% of the striatal DA and ~50% of the nigral DAergic neurons are lost (29). The pathogenesis of PD is currently unknown, but both environmental and genetic factors have been implicated in the neurodegenerative process leading to neuron death (22). Recently, it has been reported that in a Japanese population homozygosity for the V66M polymorphism of the BDNF gene occurs more frequently in patients with PD than in unaffected controls (30). Moreover, two single nucleotide polymorphisms at position C270T of the BDNF gene have been identified in patients with familial PD, suggesting that BDNF may play a role in the development of familial PD (31). Furthermore, it has been suggested that pathogenic αsynuclein mutations are linked to a reduced production of the trkB ligand BDNF (12). It has been speculated that deficits in neurotrophin signaling resulting from a decrease in neurotrophin levels may contribute to the death of midbrain DAergic cells in PD (9, 10). In humans, a majority of DAergic neurons of the SNpc reveal immunoreactivities for trkB and trkC (3). Therefore, Page 8 of 19 (page number not for citation purposes) reduced levels of these receptors may possibly be implicated in the degeneration of nigrostriatal DAergic neurons. Our data support the notion that the observed significant losses of DAergic neurons in the SNpc and striatal DAergic fibers may result from deficits in trkB- and trkC-mediated signaling. Whether the observed phenotype results exclusively from trkB and trkC deficits of nigral DAergic neurons or might imply further neuronal systems employing these receptors will have to be resolved by analyzing mouse mutants with conditional deletions of trkB and trkC genes in the DAergic system. Abnormal accumulation of α-synuclein in the soma of nigral neurons A further hallmark of PD is the abnormal accumulation of α-synuclein in cell bodies of the SN (23). Not only in human PD, but also in the MPTP mouse model of PD, there is a reduction in DAergic functions and an abnormal accumulation of α-synuclein in neurons in the SNpc (32, 33). Moreover, overexpression of α-synuclein in mice results in a progressive accumulation of αsynuclein-ir inclusions in neurons of the SNpc and a loss of DAergic terminals in the striatum (34). Targeted overexpression of α-synuclein in the rat nigrostriatal system has been shown to induce cellular and axonal pathologies, including α-synuclein-positive cytoplasmic inclusions similar to those seen in brains of PD patients (35). These changes are accompanied by a loss of 30-80% of the nigral DAergic neurons and a 10-50% reduction of TH-ir fibers in the striatum (35). However, significant motor impairment only developed in those animals in which DAergic neuron loss exceeded a critical threshold of 50-60% (35). The examined aged heterozygous trkB, trkC, and trkB/C knockout mice displayed no obvious motor impairment (data not shown), probably since TH-ir neurons were not decreased by >50%. In support of the view that the phenotype of the trkB and trkC heterozygous mutant mice may be related to PD, we have shown an abnormal accumulation of α-synuclein in neurons. However, we did not find Lewy bodies. Therefore, it is conceivable that the trkB and trkC heterozygous mutant phenotypes resemble a presymptomatic phase of PD, a stage corresponding to the Braak’s stage III of idiopathic PD (36), which is characterized by the involvement of the SNpc but a lack of Lewy bodies in this area. This assumption is also based on the findings of Wakabayashi and et al. (20), who demonstrated that α-synuclein accumulation of the type 1 is found in incidental Lewy bodies disease, which is considered to be a presymptomatic phase of PD (20). Aggregates of α-synuclein in Lewy bodies and α-synuclein containing inclusions can be recognized by thioflavin S, while small punctate aggregates are thioflavin S negative, indicating a nonfibrillar conformation (37). Because we were unable to detect thioflavin S positive inclusions in the aged heterozygous trk mutants, the α-synuclein positive aggregates in the SNpc do not represent fibrillar inclusions, strengthening the view that the morphological phenotype of the aged mutant mice resembles a presymptomatic phase. It has been suggested that the small spherical aggregates are formed before the formation of fibrillar inclusions and that they are the cellular equivalents of protofibrils (37). Furthermore, it seems that the protofibril, rather than the fibril itself, may be pathogenic (38). Even in PDdiseased brains, the majority of SNpc neurons undergoing apoptotic-like cell death does not Page 9 of 19 (page number not for citation purposes) appear to contain somal LBs and thus may be dying before LB formation can occur (39). Therefore, it has been speculated that the inclusion bodies might have a protective role and that the inclusions may sequester toxic species and/or divert α-synuclein from toxic assembly pathways (38). The proteins that have not been sequestered in the inclusion bodies may lead to cell death (40). Rotenone, paraquate, and the neurotoxic ion MPP+ (generated from MPTP) as well as 6-OHDA are toxic to the mitochondrial complex I, and application of these neurotoxins reproduces specific features of PD in animal models (22). Exposure of cells to mitochondrial inhibitors (e.g., rotenone) produces two types of α-synuclein aggregates, small punctate aggregate, and large inclusion bodies (14). Interestingly, it has been shown that BDNF increases rat brain mitochondrial respiratory coupling at complex I (41). These findings indicate that neurotrophic support can change the efficacy of coupling at complex I. Reduced signaling of neurotrophins may induce mitochondrial dysfunction associated with complex I and may lead, by yet to be discovered pathways, to the formation of small punctate α-synuclein aggregates. CONCLUSION We have discovered a phenotype of the nigrostriatal system in aged heterozygous trkB and/or trkC mutant mice that resembles a preclinical stage of PD and includes losses of TH-positive neurons in the SNpc, which are accompanied by losses of striatal TH-positive fibers and massive deposits of α-synuclein in DAergic cell bodies. These data substantiate the view that neurotrophin signaling is physiologically relevant for the maintenance of the DAergic nigrostriatal system and that its deterioration may contribute to elicit a PD-like phenotype. ACKNOWLEDGMENTS This study was supported by the DFG Forschergruppe FOR 302/TP A1 and SFB 636/A5. We wish to thank Ulla Hinz for excellent technical assistance. REFERENCES 1. Altar, C. A., Siuciak, J. A., Wright, P., Ip, N. Y., Lindsay, R. M., and Wiegand, S. J. (1994) In situ hybridization of trkB and trkC receptor mRNA in rat forebrain and association with high-affinity binding of [125I]BDNF, [125I]NT-4/5 and [125I]NT-3. Eur. J. Neurosci. 6, 1389–1405 2. Katoh-Semba, R., Semba, R., Takeuchi, I. K., and Kato, K. (1998) Age-related changes in levels of brain-derived neurotrophic factor in selected brain regions of rats, normal mice and senescence-accelerated mice: a comparison to those of nerve growth factor and neurotrophin-3. Neurosci. Res. 31, 227–234 3. Nishio, T., Furukawa, S., Akiguchi, I., and Sunohara, N. (1998) Medial nigral dopamine neurons have rich neurotrophin support in humans. Neuroreport 9, 2847–2851 4. Hagg, T. (1998) Neurotrophins prevent death and differentially affect tyrosine hydroxylase of adult rat nigrostriatal neurons in vivo. Exp. Neurol. 149, 183–192 Page 10 of 19 (page number not for citation purposes) 5. Venero, J. L., Beck, K. D., and Hefti, F. (1994) 6-Hydroxydopamine lesions reduce BDNF mRNA levels in adult rat brain substantia nigra. Neuroreport 5, 429–432 6. Levivier, M., Przedborski, S., Bencsics, C., and Kang, U. J. (1995) Intrastriatal implantation of fibroblasts genetically engineered to produce brain-derived neurotrophic factor prevents degeneration of dopaminergic neurons in a rat model of Parkinson’s disease. J. Neurosci. 15, 7810–7820 7. Klein, R. L., Lewis, M. H., Muzyczka, N., and Meyer, E. M. (1999) Prevention of 6hydroxydopamine-induced rotational behavior by BDNF somatic gene transfer. Brain Res. 847, 314–320 8. Galpern, W. R., Frim, D. M., Tatter, S. B., Altar, C. A., Beal, M. F., and Isacson, O. (1996) Cell-mediated delivery of brain-derived neurotrophic factor enhances dopamine levels in an MPP+ rat model of substantia nigra degeneration. Cell Transplant. 5, 225–232 9. Howells, D. W., Porritt, M. J., Wong, J. Y., Batchelor, P. E., Kalnins, R., Hughes, A. J., and Donnan, G. A. (2000) Reduced BDNF mRNA expression in the Parkinson’s disease substantia nigra. Exp. Neurol. 166, 127–135 10. Mogi, M., Togari, A., Kondo, T., Mizuno, Y., Komure, O., Kuno, S., Ichinose, H., and Nagatsu, T. (1999) Brain-derived growth factor and nerve growth factor concentrations are decreased in the substantia nigra in Parkinson’s disease. Neurosci. Lett. 270, 45–48 11. Parain, K., Murer, M. G., Yan, Q., Faucheux, B., Agid, Y., Hirsch, E., and Raisman-Vozari, R. (1999) Reduced expression of brain-derived neurotrophic factor protein in Parkinson’s disease substantia nigra. Neuroreport 10, 557–561 12. Kohno, R., Sawada, H., Kawamoto, Y., Uemura, K., Shibasaki, H., and Shimohama, S. (2004) BDNF is induced by wild-type alpha-synuclein but not by the two mutants, A30P or A53T, in glioma cell line. Biochem. Biophys. Res. Commun. 318, 113–118 13. Minichiello, L., and Klein, R. (1996) TrkB and TrkC neurotrophin receptors cooperate in promoting survival of hippocampal and cerebellar granule neurons. Genes Dev. 10, 2849– 2858 14. Lee, H. J., Shin, S. Y., Choi, C., Lee, Y. H., and Lee, S. J. (2002) Formation and removal of alpha-synuclein aggregates in cells exposed to mitochondrial inhibitors. J. Biol. Chem. 277, 5411–5417 15. West, M. J., Slomianka, L., and Gundersen, H. J. (1991) Unbiased stereological estimation of the total number of neurons in the subdivisions of the rat hippocampus using the optical fractionator. Anat. Rec. 231, 482–497 16. Abercrombie, M. (1946) Estimation of nuclear population from microtome sections. Anat. Rec. 94, 239–247 Page 11 of 19 (page number not for citation purposes) 17. von Bartheld, C. S. (2002) Counting particles in tissue sections: choices of methods and importance of calibration to minimize biases. Histol. Histopathol. 17, 639–648 18. Eilam, R., Peter, Y., Elson, A., Rotman, G., Shiloh, Y., Groner, Y., and Segal, M. (1998) Selective loss of dopaminergic nigro-striatal neurons in brains of Atm-deficient mice. Proc. Natl. Acad. Sci. USA 95, 12653–12656 19. von Bohlen und Halbach, O., Unsicker, K. (2003) Age-related decline in the tyrosine hydroxylase-immunoreactive innervation of the amygdala and dentate gyrus in mice. Cell Tissue Res. 311, 139-143 20. Wakabayashi, K., Hayashi, S., Kakita, A., Yamada, M., Toyoshima, Y., Yoshimoto, M., and Takahashi, H. (1998) Accumulation of alpha-synuclein/NACP is a cytopathological feature common to Lewy body disease and multiple system atrophy. Acta Neuropathol. (Berl.) 96, 445–452 21. Naoi, M., and Maruyama, W. (1999) Cell death of dopamine neurons in aging and Parkinson’s disease. Mech. Ageing Dev. 111, 175–188 22. von Bohlen und Halbach, O., Schober, A., Krieglstein, K. (2004) Genes, proteins, and neurotoxins involved in Parkinson’s disease. Prog. Neurobiol. 73, 151-177 23. Lotharius, J., and Brundin, P. (2002) Pathogenesis of Parkinson’s disease: dopamine, vesicles and alpha-synuclein. Nat. Rev. Neurosci. 3, 932–942 24. Emborg, M. E., Ma, S. Y., Mufson, E. J., Levey, A. I., Taylor, M. D., Brown, W. D., Holden, J. E., and Kordower, J. H. (1998) Age-related declines in nigral neuronal function correlate with motor impairments in rhesus monkeys. J. Comp. Neurol. 401, 253–265 25. Tatton, W. G., Greenwood, C. E., Verrier, M. C., Holland, D. P., Kwan, M. M., and Biddle, F. E. (1991) Different rates of age-related loss for four murine monoaminergic neuronal populations. Neurobiol. Aging 12, 543–556 26. Masuoka, D. T., Jonsson, G., and Finch, C. E. (1979) Aging and unusual catecholaminecontaining structures in the mouse brain. Brain Res. 169, 335–341 27. Gozlan, H., Daval, G., Verge, D., Spampinato, U., Fattaccini, C. M., Gallissot, M. C., El Mestikawy, S., and Hamon, M. (1990) Aging associated changes in serotoninergic and dopaminergic pre- and postsynaptic neurochemical markers in the rat brain. Neurobiol. Aging 11, 437–449 28. Miguez, J. M., Aldegunde, M., Paz-Valinas, L., Recio, J., and Sanchez-Barcelo, E. (1999) Selective changes in the contents of noradrenaline, dopamine and serotonin in rat brain areas during aging. J. Neural Transm. 106, 1089–1098 29. Dunnett, S. B., and Bjorklund, A. (1999) Prospects for new restorative and neuroprotective treatments in Parkinson’s disease. Nature 399, A32–A39 Page 12 of 19 (page number not for citation purposes) 30. Momose, Y., Murata, M., Kobayashi, K., Tachikawa, M., Nakabayashi, Y., Kanazawa, I., and Toda, T. (2002) Association studies of multiple candidate genes for Parkinson’s disease using single nucleotide polymorphisms. Ann. Neurol. 51, 133–136 31. Parsian, A., Sinha, R., Racette, B., Zhao, J. H., and Perlmutter, J. S. (2004) Association of a variation in the promoter region of the brain-derived neurotrophic factor gene with familial Parkinson’s disease. Parkinsonism Relat. Disord. 10, 213–219 32. Dauer, W., and Przedborski, S. (2003) Parkinson’s disease: mechanisms and models. Neuron 39, 889–909 33. Hirsch, E. C., Hoglinger, G., Rousselet, E., Breidert, T., Parain, K., Feger, J., Ruberg, M., Prigent, A., Cohen-Salmon, C., and Launay, J. M. (2003) Animal models of Parkinson’s disease in rodents induced by toxins: an update. J. Neural Transm. Suppl. 65, 89–100 34. Masliah, E., Rockenstein, E., Veinbergs, I., Mallory, M., Hashimoto, M., Takeda, A., Sagara, Y., Sisk, A., and Mucke, L. (2000) Dopaminergic loss and inclusion body formation in alpha-synuclein mice: implications for neurodegenerative disorders. Science 287, 1265– 1269 35. Kirik, D., Rosenblad, C., Burger, C., Lundberg, C., Johansen, T. E., Muzyczka, N., Mandel, R. J., and Bjorklund, A. (2002) Parkinson-like neurodegeneration induced by targeted overexpression of alpha-synuclein in the nigrostriatal system. J. Neurosci. 22, 2780–2791 36. Braak, H., Del Tredici, K., Bratzke, H., Hamm-Clement, J., Sandmann-Keil, D., and Rub, U. (2002) Staging of the intracerebral inclusion body pathology associated with idiopathic Parkinson’s disease (preclinical and clinical stages). J. Neurol. 249, Suppl. 3, 1–5 37. Lee, H. J., and Lee, S. J. (2002) Characterization of cytoplasmic alpha-synuclein aggregates. Fibril formation is tightly linked to the inclusion-forming process in cells. J. Biol. Chem. 277, 48976–48983 38. Goldberg, M. S., and Lansbury, P. T., Jr. (2000) Is there a cause-and-effect relationship between alpha-synuclein fibrillization and Parkinson’s disease? Nat. Cell Biol. 2, E115– E119 39. Tompkins, M. M., and Hill, W. D. (1997) Contribution of somal Lewy bodies to neuronal death. Brain Res. 775, 24–29 40. Ciechanover, A. (2001) Linking ubiquitin, parkin and synphilin-1. Nat. Med. 7, 1108–1109 41. Markham, A., Cameron, I., Franklin, P., and Spedding, M. (2004) BDNF increases rat brain mitochondrial respiratory coupling at complex I, but not complex II. Eur. J. Neurosci. 20, 1189–1196 Received February 23, 2005; accepted June 8, 2005. Page 13 of 19 (page number not for citation purposes) Table 1 Total neuronal numbers in the SNpc, estimated by using Nissl stained material n Mean (N)* SD CE BV Adult wt 3 9712.7 595.22 0.035 0.061 Aged Wt 3 8399.3 302.99 0.021 0.036 Aged trkB(+/−) 3 7143.0 198.01 0.016 0.028 Aged trkC(+/−) 3 7605.7 114.73 0.009 0.015 Aged trkB/C(+/−)/(+/−) 3 6613.6 128.27 0.011 0.019 *Mean values for total neuronal numbers (N); n = animals used. Page 14 of 19 (page number not for citation purposes) Table 2 Numbers of DAergic cells in the SNpc, estimated by using anti-TH stained material n Mean (N)* SD CE BV Adult wt 4 6674.5 266.727 0.020 0.039 Aged wt 3 5454.0 151.21 0.016 0.028 Adult trkB(+/−) 4 6212.5 82.68 0.007 0.013 Aged trkB(+/−) 4 4482.0 377.09 0.049 0.084 Adult trkC(+/−) 3 6276.4 403.18 0.033 0.064 Aged trkC(+/−) 4 4665.3 258.39 0.032 0.055 Adult trkB/C Aged trkB/C (+/−)/(+/−) (+/−)/(+/−) 3 6165.3 83.54 0.008 0.013 3 4034.4 326.89 0.047 0.081 *Mean values for total numbers of stained cells (N); n = animals used. Page 15 of 19 (page number not for citation purposes) Fig. 1 Figure 1. Haploinsufficiency for trkB, trkC or trkB/C induces alterations in the SNpc and striatum. A) Neuronal numbers in the SNpc. B) Numbers of TH-immunopositive neurons in the SNpc of adult mice. C) Numbers of TH-immunopositive neurons in the SNpc of aged mice. D) Relative densities of striatal TH-ir fibers of adult mice. E) Relative densities of striatal TH-ir fibers of aged mice. (*, compared with adult wild-type littermates; ♦, compared with aged wild-type littermates; one symbol indicates P<0.05; two symbols indicate P<0.01; error bars display SD). Page 16 of 19 (page number not for citation purposes) Fig. 2 Figure 2. Densities of cells in the substantia nigra with α-synuclein positive aggregates within the soma of adult (A) and aged (B) mice. ANOVA was used to examine whether changes in the density of α-synuclein accumulating cells were significant (*, compared with adult mice; ♦, compared with aged wild-type littermates; one symbol indicates P<0.05; two symbols indicate P<0.01; error bars display SE). Page 17 of 19 (page number not for citation purposes) Fig. 3 Figure 3. Alpha-synuclein accumulating neurons in the SNpc. A-C) Alpha-synuclein accumulating neurons in the SNpc of an aged trkB(+/−) mouse (A), of an aged trkC(+/−) mouse (B), and of an aged trkB/trkC(+/−)/(+/−) mouse (C). D–F) Neurons in the SNpc of aged heterozygous trkB/trkC(+/−)/(+/−) mice are immunoreactive for α-synuclein (D: red signal) and TH (E: green). The overlay (F) shows that α-synuclein is accumulated in DAergic neurons of the SNpc. G-I) Alpha-synuclein immunoreactivity (G, in red) and TH immunoreactivity (H, in green) are predominantly colocalized. Only very rare αsynuclein accumulating cells were found (I), which were not TH-positive [marked with an arrow; cell nuclei were counterstained with DAPI (in blue)]. J–L) Neurons in the SNpc of aged wild-type mice. No abnormal accumulation of α-synuclein can be found (J). K shows TH-ir in the SNpc, and L is an overlay of J and K [cell nuclei were counter-stained with DAPI (in blue)]. Page 18 of 19 (page number not for citation purposes) Fig. 4 Figure 4. No signs of amyloid-like inclusions in the SNpc of aged heterozygous trkB/trkC(+/−)/(+/−) mice. Neurons in the SNpc of aged heterozygous trkB/trkC(+/−)/(+/−) mice are immunoreactive for TH (A: red signal), but they are thioflavin S negative (B). The overlay is presented in C [cell nuclei were counter-stained with DAPI (in blue)]. Page 19 of 19 (page number not for citation purposes)