Introduction
Parkinson’s disease (PD) is the second most common neurodegenerative disease after Alzheimer’s disease, and Lewy bodies (LBs) and Lewy neurites (LNs) are characteristic features of PD. Dementia with Lewy bodies (DLB) is also a progressive neurodegenerative disease characterized by the appearance of LBs and LNs in cortex [
17,
22]. The discovery of disease-associated mutation in the α-synuclein gene
SNCA and subsequent immunostaining studies with antibodies demonstrated that α-synuclein is the major component of LBs and LNs [
2,
55,
56]. It is also the major component of glial cytoplasmic inclusions (GCIs) in multiple system atrophy (MSA) [
54,
58]. These diseases are collectively referred to as α-synucleinopathies. To date, six missense mutations in the
SNCA gene and occurrence of gene multiplication have been identified in familial forms of PD and DLB [
1,
5,
24,
28,
29,
41,
52,
62]. α-Synuclein is a small protein of 140 amino acids, which is localized in presynaptic termini, and is involved in maintenance of synapses and synaptic plasticity. In PD, DLB, or MSA patients, it is deposited in the brain as a filamentous form with cross-β structure [
51], which is abnormally phosphorylated at Ser129 and partially ubiquitinated [
15,
21]. α-Synuclein is natively unfolded, but readily assembles into amyloid-like fibrils under appropriate conditions. Pathogenic mutations affect fibril formation in vitro, either accelerating fibril formation [
6,
7,
16] or resulting in formation of fibrils that are more fragile and easier to propagate than wild-type (WT) fibrils [
61]. Moreover, the spreading of pathological α-synuclein is closely correlated with disease progression; indeed, the distribution pattern and spread of the pathologies are useful for disease staging of sporadic PD [
3,
48]. These results suggest that intracellular amyloid-like α-synuclein fibrils can cause PD and DLB, and spreading of α-synuclein pathology in the brain is considered to be the underlying mechanism of progression of these diseases. Recently, it was experimentally demonstrated that intracerebral injection of synthetic α-synuclein fibrils and/or insoluble α-synuclein from diseased brain converts normal α-synuclein into an abnormal form, and the abnormal α-synuclein propagates throughout the brain in a prion-like manner in WT mouse [
30,
33,
34,
57], α-synuclein transgenic mouse [
31,
36,
60] and monkey [
44].
Common marmoset (
Callithrix jacchus) is a very small new world primate, about 25 – 35 cm in height and 300 – 500 g in weight, and is far more experimentally tractable than macaque monkey. Since it has high fecundity, with a short sexual maturation period of 18 months, it is attracting increasing attention as an experimental model of primates. In fact, a national project called Brain/MINDS (Brain Mapping by Integrated Neurotechnologies for Disease Studies) was started in 2014 in Japan to develop the common marmoset as a model animal for neuroscience [
19,
38,
39]. The marmoset cortex is relatively smooth, but the gyrencephalic and cortical sheet is divided into functionally distinct cortical areas, as in Old World monkeys [
45], and thus is suitable for studies of higher cognitive functions and social communication [
11]. Therefore, marmosets are considered to be a good experimental model animal to understand the evolution of brain development and function. Moreover, transgenic marmosets have already been generated, demonstrating the feasibility of gene manipulation in this species [
49].
To date, mouse models have been used to investigate brain development, circuits, and higher cognitive functions, but they have limitations for exploration of the evolution and development of the primate neocortex. In situ hybridization analysis of marmoset brain revealed that the expression patterns of the genes that regulate brain development (such as EphA6) are different, especially in brain areas that have connections to the prefrontal cortex and are presumably involved in higher cognitive functions, although similar broad regional patterns of expression were observed in both species [
32].
A particular difference in brain development and structure between mouse and marmoset is that striatum of marmoset is separated into caudate nucleus and putamen, while these are not distinguishable in rodents. It has been considered that caudate nucleus and putamen were originally one structure and that they became separated by the internal capsule during evolution [
25]. Thus, the marmoset has advantageous characteristics as an experimental animal to study brain networks, functions and disease conditions.
Here, we investigated whether intracerebral injection of α-synuclein fibrils can induce PD/DLB-like pathologies in marmoset, and we present the first marmoset model of α-synuclein propagation. We found that marmosets developed abundant phosphorylated α-synuclein pathologies, similar to those observed in PD/DLB, in various brain regions, including striatum, cortex and substantia nigra, at only three months after injection. Remarkably, many LB-like inclusions are observed in tyrosine hydroxylase (TH)-positive dopamine neurons, and a significant decrease in TH-staining was seen in the injection hemisphere. The inclusions were also positive for fluorescent β-sheet ligands, thioflavin-S and FSB, implying that α-synuclein deposits in these animals should be detectable in vivo by positron emission tomography (PET) with a suitable small-molecular agent. Taking account of the advantages of marmosets over mice, we believe the current experimental model would be particularly useful to examine the relationships between PET-detectable α-synuclein lesions and disruptions of neural networks in the absence and presence of candidate α-synucleinopathy-modifying therapeutics.
Materials and methods
Preparation of recombinant α-synuclein and fibrils
Recombinant human and mouse wild-type α-synuclein and fibrils were prepared as described previously [
33,
57]. Briefly, purified α-synuclein (7 – 10 mg/ml) was incubated at 37 °C in a shaking incubator at 200 rpm in 30 mM Tris–HCl, pH 7.5, containing 0.1% NaN
3, for 72 h. α-Synuclein fibrils were pelleted by spinning the assembly mixtures at 113,000 xg for 20 min, resuspended in 30 mM Tris–HCl buffer (pH 7.5), and sonicated for 3 min (Biomic 7040 Ultrasonic Processor, Seiko). The protein concentrations were determined by HPLC. Samples were run on gradient 12% polyacrylamide gels and stained with Coomassie Brilliant Blue (CBB), or electrophoretically transferred to PVDF membranes. For immunoblotting, membranes were incubated with 3% gelatin (Wako) for 10 min at 37 °C, followed by overnight incubation at room temperature with primary antibodies. Next, the membranes were incubated for 1 hr at room temperature with biotinylated anti-rabbit or mouse IgG (Vector Lab), then incubated for 30 min with avidin-horseradish peroxidase (Vector Lab), and the reaction product was visualized by using 0.1% 3,3-diaminobenzidine (DAB) and 0.2 mg/ml NiCl
2 as the chromogen. For electron microscopy, samples were placed on collodion-coated 300-mesh copper grids, stained with 2% (v/v) phosphotungstate, and examined with a JEOL 1200EX electron microscope.
Marmosets
According to animal protection considerations based on the 3R (reduce, reuse, recycle) principle, we designed the experiment very carefully to minimize the number of animals used. Two female 26-month-old marmosets (individual recognition No. 14H and 14I; born on 4th April, 2014 and bred at the Animal Research Division, Tokyo Metropolitan Institute of Medical Science) were used for this experiment.
Stereotaxic surgery
The marmosets were anesthetized with Ketamine Hydrochloride (20–40 mg/kg i.m.) and Xylazine (0.05 mg/kg i.m.), and Butorphanol (0.05–0.1 mg/kg i.m.). Then, 50 μL aliquots of 4 mg/mL mouse α-synuclein fibrils were injected into both caudate nucleus (interaural +9.5 mm, Lateral 3 mm, Depth 6 mm) and putamen (interaural +9.5 mm, Lateral 6 mm, Depth 3 mm) in the right hemisphere of 14H brain (total 400 μg). A 50 μL aliquot was injected into caudate nucleus (interaural +9.5 mm, Lateral 3 mm, Depth 6 mm) in the right hemisphere of 14I brain (total 200 μg). The marmosets were bred for 3 months after injection in a biological safety level 2 (BSL-2) environment. All experimental protocols were approved by the Animal Care and Use Committee of Tokyo Metropolitan Institute of Medical Science (No. 16038).
Antibodies
Primary antibodies used in this study are listed in Table
1. An anti-phosphorylated α-synuclein rabbit monoclonal antibody to pS129 (Abcam) and other anti-α-synuclein antibodies, including LB509 [
26] (a gift from Dr Iwatsubo), 75–91 (Cosmo bio), 131–140 (Cosmo bio) and #2642 (Cell Signaling Technology) were used for detection and characterization of α-synuclein pathologies in marmoset brains. Anti-p62 (Progen), anti-Ub (Dako, Millipore), anti-TH (Millipore), anti-NeuN (Millipore), anti-GFAP (Sigma), anti-CNPase (Abcam) and anti-Iba1 (Wako) antibodies were also used.
Table 1
Antibodies used in this study
pS129 (phosphorylated a-syn) | rabbit mono | Abcam (ab51253) | 1:2000 |
LB509 (human a-syn) | mouse mono | Gift from Dr Iwatsubo | 1:1000 |
75–91 (a-syn 75–91) | rabbit poly | Cosmo bio (CAC-TIPSNP08) | 1:1000 |
131–141 (a-syn 131–140) | rabbit poly | Cosmo bio (CAC-TIPSNP09) | 1:1000 |
#2642 (a-syn) | rabbit poly | Cell Signaling Tech (#2642) | 1:1000 |
Anti-p62 | guinea pig poly | Progen (GP62-C) | 1:1000 |
Anti-Ub | rabbit poly | Dako (Z0458) | 1:1000 |
Anti-Ub | mouse mono | Millipore (MAB1510) | 1:1000 |
Anti-TH | rabbit poly | Millipore (AB152) | 1:1000 |
Anti-TH | mouse mono | Millipore (MAB318) | 1:1000 |
Anti-NeuN | mouse mono | Millipore (MAB377) | 1:1000 |
Anti-GFAP | mouse mono | Sigma (G3893) | 1:1000 |
Anti-CNPase | mouse mono | Abcam (ab6319) | 1:200 |
Anti-Iba1 | rabbit poly | Wako (016–20001) | 1:1000 |
Immunohistochemistry
Marmosets were deeply anesthetized with pentobarbital injection and killed, and the brain was perfused with 0.1 M phosphate buffer, followed by 10% formalin neutral buffer solution. After fixation, whole brains were sectioned coronally at 50 μm using a vibratome (Leica, Wetzlar, Germany). For high-sensitivity detection, free-floating brain sections were treated with formic acid for 20 min, washed, and boiled at 100 °C for 20 min as described [
33]. Sections were then incubated with 0.5% H
2O
2 in methanol for 30 min to inactivate endogenous peroxidases, blocked with 10% calf serum in PBS for 20 min, and incubated overnight with appropriate antibodies. After incubation with the biotinylated secondary antibody for 2 h, labeling was detected using the ABC staining kit (Vector) with DAB. Sections were counterstained with hematoxylin. Slides were coverslipped with mounting medium. Images were observed with an all-in-one microscope/digital camera (BZ-X710; Keyence).
For double-label immunofluorescence detection, brain sections were pretreated as described above and incubated overnight at 4 °C with a cocktail of appropriate primary antibodies. The sections were washed and incubated with a cocktail of Alexa568-conjugated goat anti-mouse or anti-rabbit IgG and Alexa488-conjugated goat anti-mouse or anti-rabbit or anti-guinea pig IgG (Molecular Probes). After further washing, the sections were coverslipped with non-fluorescent mounting media (VECTASHIELD; Vector Laboratories) and observed with the BZ-X710.
To measure positive cells, 7–9 sections of substantia nigra were randomly selected, and all images were captured with BZ-X710 microscope using the same settings. The areas of pS129-positive cells and TH-positive cells in the right and left substantia nigra were extracted and quantified by BZ-H3C Hybrid Cell Count Software (Keyence).
Thioflavin-S and FSB stainings
Thioflavin-S and FSB [
23,
50] were purchased from Sigma-Algrich and Dojindo, respectively. For fluorescence labeling with β-sheet ligands, thioflavin-S and FSB, brain sections were mounted on a glass slide and dried with warm air. Sections were incubated in 20% ethanol containing 0.001% β-sheet ligands at room temperature for 30 min. The samples were rinsed with 20% ethanol for 5 min, dipped into distilled water twice for 3 min each, and mounted in VECTASHIELD. Fluorescence images were captured using a BZ-X710 fluorescence microscope equipped with Filter set ET-ECFP (Chroma Technology) for thioflavin-S and FSB. After fluorescence microscopy, all sections labeled with ligands were autoclaved for antigen retrieval, immunostained with the pS129 antibody, and examined using the BZ-X710 instrument.
Discussion
Emerging evidence indicates that intracellular amyloid-like proteins have prion-like properties and propagate from cell to cell by converting normal proteins into abnormal forms [
18,
22,
27,
42]. This prion-like propagation may account for the characteristic spreading of pathological proteins including α-synuclein, tau and TDP-43, and also for disease progression in major neurodegenerative diseases with these protein pathologies, such as Parkinson’s disease, Alzheimer’s disease and amyotrophic lateral sclerosis.
In this study, we tested whether inoculation of synthetic mouse α-synuclein fibrils can induce PD-like α-synuclein pathologies and prion-like propagation in two adult marmosets by injecting the fibrils into striatum (one animal was injected into caudate nucleus and the other, into caudate nucleus and putamen). Within only 3 months after injection, we observed abundant phospho-α-synuclein pathologies in various brain regions of both marmosets, indicating that prion-like conversion readily occurred in the primate brains even within this short time-scale. Luk et al. and we have established a propagation model in wild-type mouse [
30,
33,
34,
57], but others have found it difficult to detect the pathologies in wild-type mouse [
46]. It has also been reported that intranigral or intrastriatal inoculations of PD-derived LB extracts in monkey resulted in progressive nigrostriatal neurodegeneration, but clearly defined LB-type inclusions were not observed [
44]. The results of the present study clearly demonstrate that inoculation of fibrillar α-synuclein in striatum of wild-type marmoset triggered PD-like α-synuclein pathologies, which propagated retrogradely to substantia nigra and other input regions, and induced degeneration of dopaminergic neurons. Furthermore, most of the inclusions were positive for amyloid-sensitive dyes, such as thioflavin-S and FSB. This simple experiment has provided direct evidence for prion-like propagation of pathological α-synuclein in brains of primates, and the model should be very useful for establishing in vivo imaging methodology for abnormal α-synuclein propagation and for development and evaluation of disease-modifying drugs for α-synucleinopathies.
In this study, we did not perform behavioral tests, but we did not observe any apparent symptoms or behavior deficits in these marmosets, suggesting that they may not develop strong phenotypes within 3 months after inoculation. It is reasonable to speculate that motor deficits would only be detected after the loss of more than 50% of dopamine neurons, as is the case in PD patients. We observed 20 – 40% decrease of TH-positive cells in the right hemisphere in the animals in this study. Further studies will be needed to establish the relationship between pathologies and symptoms in wild-type marmosets.
The present model should be useful for research on PD and α-synucleinopathies, because this is a primate and non-transgenic wild-type animal model, which would not suffer from various artifacts associated with overexpression of proteins in transgenic animals [
47] or the use of viral vector-mediated gene transfer systems. Among mouse models, Tg-mice overexpressing human A53T mutant α-synuclein (such as M83 line) are considered a good host animal for inoculation experiments, because disease symptoms and α-synuclein pathologies appear at about ~100 days after inoculation [
43]. When Tg-marmoset models overexpressing human α-synuclein are available, it will be interesting to inject synthetic α-synuclein fibrils or brain extracts from patients into these animals to see whether the appearance of PD-like symptoms or pathologies is accelerated.
By double immunolabeling of marmoset brain sections with LB509 and Iba1, we demonstrated that some of the α-synuclein inclusions are colocalized with Iba1-positive microglial cells. This finding suggests that inclusions or degenerating neurons with aggregates may be phagocytosed by microglial cells. Although it has been debated whether inflammation constitutes a cause or consequence of PD, increasing evidence suggests that microglial cells and inflammatory pathways are involved in the pathogenesis and progression of PD [
8,
9]. Indeed, activated microglia are prevalent in the most pathologically affected areas in the brains of PD patients [
9,
35]. Recent studies also demonstrated that toll-like receptor 2 may contribute to α-synuclein pathology in PD [
10]. However, there has been no direct evidence that microglial cells are involved in the clearance of α-synuclein aggregates, and our findings here represent the evidence that α-synuclein aggregates or cells with inclusions are phagocytosed by microglial cells for clearance. It seems plausible that such microglial phagocytosis of α-synuclein inclusions may be a protective event to clear degenerating neurons and reduce inflammation in the brain, but further studies will be needed to confirm this. Our marmoset model should be useful for elucidating the molecular mechanisms of α-synuclein propagation, and also for exploring neuronal circuits in marmoset brain and human brain.
Acknowledgments
This work was supported by Ministry of Education, Culture, Sports, Science, and Technology Grants-in-Aid for Scientific Research (KAKENHI) Grants JP26117005 (to M. H.), Japan Society for the Promotion of Science Grants-in-Aid for Scientific Research (KAKENHI) Grant JP23228004 (to M. H.), and a grant-in-aid for research on Brain Mapping by Integrated Neurotechnologies for Disease Studies (Brain/MINDS) from the Japan Agency for Medical Research and Development (AMED) JP14533254 (to M. H. and M. H.). The authors declare that they have no conflicts of interest with the contents of this article.