Dopaminergic Loss and Inclusion Body Formation in α-Synuclein Mice: Implications for Neurodegenerative Disorders

See allHide authors and affiliations

Science  18 Feb 2000:
Vol. 287, Issue 5456, pp. 1265-1269
DOI: 10.1126/science.287.5456.1265


To elucidate the role of the synaptic protein α-synuclein in neurodegenerative disorders, transgenic mice expressing wild-type human α-synuclein were generated. Neuronal expression of human α-synuclein resulted in progressive accumulation of α-synuclein—and ubiquitin-immunoreactive inclusions in neurons in the neocortex, hippocampus, and substantia nigra. Ultrastructural analysis revealed both electron-dense intranuclear deposits and cytoplasmic inclusions. These alterations were associated with loss of dopaminergic terminals in the basal ganglia and with motor impairments. These results suggest that accumulation of wild-type α-synuclein may play a causal role in Parkinson's disease and related conditions.

Human α-synuclein is a 140–amino acid molecule (1) that is encoded by a gene on chromosome 4 (2) and has homology to rat and Torpedo α-synuclein and songbird synelfin (3). Although the precise function of the synuclein superfamily of peptides is still unknown, several lines of evidence suggest potential roles in synaptic function and neural plasticity (3, 4). Human α-synuclein was originally isolated from plaques of Alzheimer's disease brains as a 19-kD protein precursor of the highly hydrophobic 35–amino acid metabolite, nonamyloid component (NAC) of plaques (1). The NAC peptide can self-aggregate into fibrils and induces aggregation of the β-amyloid peptide (5). α-Synuclein is highly abundant in presynaptic terminals (4) and in Lewy bodies (6), neuronal inclusions that are found in diverse human neurodegenerative disorders including the Lewy body variant of Alzheimer's disease, diffuse Lewy body disease, and Parkinson's disease (7). Rare cases of familial Parkinson's disease have recently been linked to point mutations in α-synuclein (8); however, most neurodegenerative disorders with Lewy bodies are associated with abnormal accumulation of wild-type, not mutant, α-synuclein (6, 9).

To elucidate the role of α-synuclein accumulation in the pathogenesis of neurodegenerative disorders with Lewy bodies, we generated and analyzed α-synuclein transgenic mice. In these mice, wild-type human α-synuclein was expressed under the regulatory control of the platelet-derived growth factor–β (PDGF-β) promoter (Fig. 1A) (10). This promoter was chosen because it has been successfully used to target the expression of other human proteins to neurons in transgenic models of neurodegenerative disease (11). In all lines of transgenic mice with human α-synuclein mRNA expression in the brain (Fig. 1, B and C), cerebral human α-synuclein expression was also detected at the protein level (Fig. 1, E and F). Mice from line D consistently had the highest levels of human α-synuclein mRNA (Fig. 1C) and protein (Fig. 1F). Lines A, B, M, and C had low to intermediate levels of transgene expression (Fig. 1, C and F).

Figure 1

Characterization of α-synuclein expression in human α-synuclein transgenic mice. (A) Human PDGF- β promoter-driven trans- gene encoding wild-type human α-synuclein. h, human; pA, polyadenylation signal. (B) Representative autoradiograph showing cerebral α- synuclein mRNA levels in transgenic (Tg) mice of different lines, a nontransgenic (Ntg) mouse, and a human (H) without neurological disease. The frontoparietal cortex was analyzed by ribonuclease protection assay as described (22) with RNA probes specific for the entire coding sequence of human α- synuclein (hα-syn, GenBank accession number L08850) or for a segment of murine α- synuclein (mα-syn, GenBank accession number AF044672). The left-most lane shows signals of undigested (U) radiolabeled RNA probes; the other lanes contained the same RNA probes plus brain RNA samples digested with ribonucleases. Protected mRNAs are shown on the right. (C) PhosphorImager (Becton-Dickinson) analysis of human α-synuclein mRNA signals (expressed as percent of human control). The highest level of transgene expression was detected in mice from line D (n = 3) and intermediate to low levels were observed in lines M, A, B, and C (n = 3 mice per line). No human α-synuclein signal was detected in nontransgenic (Non tg) mice (n = 4). Bars represent means ± SEM. (D) Schematic representation of the α-synuclein region from which the human- and mouse-specific peptides were selected for generation of polyclonal antibodies. Mismatched amino acids are labeled with an asterisk. These antibodies were generated by Research Genetics (Huntsville, AL) as described (4) and affinity-purified with the AminoLink Kit (Pierce, Rockford, IL) with the immunogen. Single-letter abbreviations for the amino acid residues are as follows: A, Ala; D, Asp; E, Glu; G, Gly; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; S, Ser; V, Val; and Y, Tyr. (E) Western blot analysis (10 μg protein per lane) with affinity-purified human- and mouse-specific α-synuclein antibodies was performed as described (4). The human-specific antibody recognized a 19-kD band, consistent with human α-synuclein, in the transgenic mice and in the human control. Only minimal cross-reactivity with mouse α-synuclein was seen in nontransgenic mice. In contrast, the murine-specific antibody recognized endogenous α-synuclein in transgenic and nontransgenic mice, with only a faint band noted in the human control. (F) PhosphorImager analysis of human α-synuclein signals (background values subtracted) from semiquantitative Western blots revealed the highest cerebral levels of human α-synuclein expression in mice from line D (n = 3) and intermediate to low levels in lines M, A, B, and C (n = 3 mice per line). The signal obtained in nontransgenic mice (n = 4) was considered background. Results were expressed as percent immunoreactivity of normal human control (n = 3). Bars represent means ± SEM.

To characterize the effects of human α-synuclein expression in neurons, we compared transgenic mice from lines A (n = 6), B (n = 8), C (n = 17), D (n = 15), and M (n = 6) with age-matched nontransgenic controls (n = 22). By 2 months of age, transgenic mice from all lines had prominent intraneuronal inclusions (nuclear and cytoplasmic) that were strongly immunoreactive with antibodies to human (Fig. 2B), but not mouse (Fig. 2E), α-synuclein. The antibody to human α-synuclein also recognized the characteristic intracytoplasmic inclusions found in Lewy body disease (Fig. 2C). Human α- synuclein−immunoreactive inclusions were most abundant in transgenic mice from the highest expresser line (Figs. 2B and3A) and were not detected in nontransgenic controls (Fig. 2A). In the transgenic mice, the inclusions were most frequently seen in neurons in the deeper layers of the neocortex (Fig. 2B), the CA3 region of the hippocampus, and the olfactory bulb (12) and occasionally in the substantia nigra (Fig. 2, G and H). These regions are also typically affected in patients with Lewy body disease (6).

Figure 2

Expression of α-synuclein immunoreactivity and neuropathological alterations in human α-synuclein transgenic mice. Brain sections from 3-month-old nontransgenic (A andD) and transgenic (B, E, andG to J) mice from line D and a human with Lewy body disease (C and F) were immunostained with an antibody to human α-synuclein (A to C and G to J) or murine α-synuclein (D to F) and imaged by light (A to G, and I) or confocal (H and J) microscopy. In nontransgenic mice, no human α-synuclein immunoreactivity was observed (A), whereas in the temporal neocortex of the transgenic mouse (B) and the human (C) there was intense α- synuclein immunoreactivity in nerve terminals and intracytoplasmic inclusions (arrows). The antibody to murine α-synuclein immunostained only the neuropil; there was no immunostaining of inclusions in the mice [(D) and (E)] or of the Lewy bodies or synapses in the human tissue (F). (G) Occasional human α-synuclein–immunoreactive structures (arrows) were observed in the cytoplasm of neurons in the substantia nigra pars compacta of transgenic mice. (H) Double labeling for human α-synuclein and TH revealed inclusion bodies (arrow) in dopaminergic neurons in the substantia nigra of the transgenic mouse. (I) Human α-synuclein–immunoreactive neuritic processes (arrows) were observed in the CA3 region of the hippocampus in transgenic mice. (J) Double labeling of transgenic brain sections revealed that some human α-synuclein–immunoreactive inclusions (red) were also ubiquitin-positive (green). Colabeled inclusions are seen in orange-yellow (arrow). Original magnification, ×295 (A to G), ×830 (H to J).

Figure 3

A high number of neuronal inclusions are associated with the development of dopaminergic and motor deficits. (A) The density of human α-synuclein–immunoreactive neuronal inclusions in the cingulate cortex was determined (13) in 3- to 4-month-old transgenic mice from lines D (n = 8), M (n = 3), A (n = 6), B (n = 4), and C (n = 7) and in age-matched nontransgenic controls (n = 8). The highest and lowest numbers of inclusion bodies were found in the highest expresser (line D) and lowest expresser (line C) transgenic lines, respectively. (B) The density of TH-positive terminals in the striatum (13) was significantly lower in 12-month-old transgenic mice from line D than in age-matched nontransgenic controls (n = 8 mice per group; *P < 0.05 by unpaired two-tailed Student's t test). (C) Western blot (23) demonstrating striatal TH levels in transgenic mice from line D and nontransgenic controls at 12 months of age. (D) Analysis of TH bands with the ImageQuant software (23) confirmed that striatal TH levels were significantly lower in transgenic mice than in nontransgenic controls (n = 5 mice per group; *P < 0.005 by unpaired two-tailed Student's t test). (E) Striatal TH activity (24) was also significantlylower in 12-month-old transgenic mice from line D than in nontransgeniclittermates (n = 5 mice per group; *P < 0.03 by unpaired two-tailed Student'st test). (F) Rotorod testing (18) revealed a significant decrease in motor performance (P < 0.01 for genotype effect by repeated-measures analysis of variance) in 12-month-old transgenic mice from line D compared with nontransgenic littermates (n = 10 mice per group). All quantitative data represent group means ± SEM.

Degeneration of dopaminergic neurons in the substantia nigra results in Parkinson's disease and is frequently preceded by the formation of Lewy bodies. To assess whether these cells develop human α-synuclein–positive inclusions in our transgenic mice, we double-immunolabeled brain sections with antibodies to human α-synuclein and to tyrosine hydroxylase (TH) (13), which is required for the synthesis of dopamine. TH-positive neurons in the substantia nigra displayed abnormal accumulations of α-synuclein (Fig. 2H). Consistent with observations in humans with Lewy body disease (6, 7), the inclusions in the transgenic mice were accompanied by occasional human α-synuclein–immunoreactive neuritic processes (Fig. 2I) and were colabeled (13) with an antibody to ubiquitin (Fig. 2J).

We then characterized the ultrastructural features of the human α-synuclein–positive neuronal inclusions by electron microscopy (4, 14). In contrast to nontransgenic controls (Fig. 4, A and B), neurons of high-expresser human α-synuclein mice (2 to 3 months of age) showed electron-dense deposits (0.1 to 0.25 μm in diameter) associated with the rough endoplasmic reticulum (Fig. 4, C and D). Larger electron-dense deposits (0.5 to 1 μm in diameter) with a dense core were observed in neuronal nuclei (Fig. 4, E and F). Interestingly, other synaptic proteins also accumulate in neuronal nuclei in other neurodegenerative conditions. For instance, fibrillar aggregates of huntingtin are found in the nucleus of affected neurons in huntingtin transgenic mice and in patients with Huntington's disease (15).

Figure 4

Ultrastructural analysis of neuronal alterations in human α-synuclein transgenic mice (line D). Upper panels show low magnification (original magnification, ×2500) and lower panels high magnification (×15,000) views of neurons from the cingulate cortex. Boxed areas in upper panels (E andG) indicate areas shown in lower panels (F andH). (A and B) Normal appearance of neuronal structures in a 3-month-old nontransgenic control. RER, rough endoplasmic reticulum. (C and D) Electron-dense inclusions in the RER in a 3-month-old transgenic mouse. (E and F) Nuclear inclusions in a 3-month-old transgenic mouse. The dense cores of these abnormal deposits distinguish them from the nuclear chromatin condensations seen in (A). (G and H) Electron-dense inclusions composed of fine granular material in the neuronal cytoplasm of an 11-month-old transgenic mouse. Immunogold electron microscopic analysis (4) of similar deposits in a different transgenic mouse (inset, H) confirmed that such electron-dense material contained human α-synuclein immunoreactivity (arrows). Bar, 5 μm.

In 9- to 11-month-old α-synuclein transgenic mice, larger electron-dense cytoplasmic inclusions (2 to 5 μm in diameter) were identified that were composed of fine granular material and contained clear vacuoles (50 to 100 nm in diameter) but no fibrillar elements (Fig. 4, G and H). Immunogold electron microscopic analysis confirmed that these structures contained human α-synuclein immunoreactivity (Fig. 4H, inset). Control experiments in which sections from transgenic mice were incubated in the absence of primary antibody showed no immunogold labeling associated with inclusions (12). Nuclear or cytoplasmic inclusions such as those depicted in Fig. 4, C to H, were not observed in nontransgenic controls or transgenic mice expressing other amyloidogenic proteins directed by the same promoter (11, 14).

The cytoplasmic neuronal inclusions in human α-synuclein transgenic mice resemble Lewy bodies in humans in some respects but not in others. Similarities include their locations in the deep layers of the neocortex and in dopaminergic neurons of the substantia nigra and their reactivity with antibodies to human α-synuclein or ubiquitin. They differ from human Lewy bodies in that they are less circumscribed, are present in the nucleus, and lack fibrillar components (6, 7). It is unclear why no fibrillar structures were detected in association with the electron-dense inclusions in transgenic mice. Conceivably, additional stress conditions, such as formation of oxygen free radicals, are necessary to promote fibrillar aggregation of human α-synuclein (16). This situation may be similar to that encountered in human amyloid protein precursor transgenic mice, where there is no formation of neurofibrillary tangles or paired helical filaments despite extensive deposition of amyloid (14).

Because motor deficits in Lewy body disease are associated with degeneration of nigral dopaminergic neurons projecting to the striatum (17), we further evaluated the integrity of the dopaminergic system. The density of TH-positive neurons in the substantia nigra was similar in all five lines of transgenic mice and in nontransgenic controls (12). However, TH-positive nerve terminals within the striatum were significantly reduced in transgenic mice from the highest expresser line as compared with nontransgenic littermates (Fig. 3B) and lower expresser lines (12). Transgenic mice also had lower striatal TH levels by Western blot analysis (Fig. 3, C and D) and lower striatal levels of TH enzymatic activity (Fig. 3E) than nontransgenic controls. Thus, abnormal accumulation of human α-synuclein may lead to injury of nerve terminals and synapses in the absence of overt neuronal loss.

To determine if loss of dopaminergic input to the striatum in transgenic mice is associated with Parkinson's disease–related neurological impairments, we examined the mice with the rotorod test (18, 19). Compared with nontransgenic littermate controls, transgenic mice from the high-expresser line D showed significant deficits in motor performance (Fig. 3F). The loss of dopaminergic terminals and motor abnormalities in human α-synuclein transgenic mice support the general hypothesis that intraneuronal accumulation of amyloidogenic synaptic proteins can elicit morphological and functional impairments of the central nervous system (20). Because the dopaminergic and behavioral deficits were detected only in the high-expresser line and not in low-expresser lines (12), these findings will need to be confirmed in additional high-expresser lines. Although the highest expresser line clearly had more human α-synuclein– and ubiquitin-immunoreactive neuronal inclusions than the other transgenic lines, these inclusions were observed in all transgenic lines analyzed (Fig. 3A). It is conceivable that a critical threshold of α- synuclein accumulation is required for dopaminergic and behavioral deficits to become detectable. Alternatively, the formation of neuronal inclusions and the development of neuronal deficits may not be linked in a causal chain. Evidence for such a dissociation has been obtained in models of polyglutamine-induced neurodegenerative disorders (21). In any case, the similarities between neurodegenerative diseases with Lewy bodies and the alterations elicited in human α- synuclein transgenic mice suggest that increased expression or intracellular accumulation of wild-type α-synuclein may play a key role in the pathogenesis of these conditions.

  • * To whom correspondence should be addressed. E-mail: emasliah{at}


Stay Connected to Science

Navigate This Article