Parkinson’s disease (PD) and dementia with Lewy bodies (DLB) are progressive neurodegenerative diseases characterized by appearance of Lewy bodies (LBs) and Lewy neurites (LNs) 1 . α-Synuclein (αsyn) is the major component of LBs and LNs, and is deposited in a hyperphosphorylated form in β-sheet-rich amyloid fibrils 2 – 5 . Five missense mutations in the αsyn gene and occurrence of gene multiplication have been identified in the familial forms of PD and DLB 6 – 13 . Moreover, it was reported that the mutations affect amyloid fibril formation in vitro, either accelerating fibril formation 14 – 16 or resulting in formation of fibrils that are more fragile and easy to propagate than wild-type (WT) fibrils 17 . These results clearly indicate that abnormalities of αsyn can induce PD and DLB. Distribution of αsyn pathology in brains with sporadic PD occurs from olfactory bulb and/or brainstem, and spreads to other brain regions concomitantly with progression of disease symptoms 18 , 19 . Thus, spread of αsyn pathology in the brain can be regarded as the underlying mechanism of progression of these diseases. Recently intracerebral injection of synthetic αsyn fibrils and/or insoluble αsyn from diseased brain was shown to induce αsyn pathology that propagated throughout the brain in a prion-like manner in WT mouse 20 , 21 , αsyn transgenic mouse 22 – 24 and monkey 25 . However, the mechanisms through which exogenous abnormal fibrils enter neurons and through which insoluble αsyn is transported to other neurons remain unknown.

To investigate where αsyn pathologies develop and how they propagate, we injected recombinant αsyn fibrils into substantia nigra, striatum, or entorhinal cortex of WT mice, and compared the spreading patterns and distribution of phosphorylated αsyn pathologies at 1 month after intracerebral injections. Our results clearly suggest that propagation of pathological αsyn occurred along neural circuits and involved trans-synaptic transport. We also showed that αsyn pathology induced tau and TDP-43 accumulation in WT mice, similar to that seen in DLB brains. This mouse model should be useful for elucidating mechanisms of disease progression of synucleinopathy and also for development of novel disease-modifying drugs.

We investigated the spread of αsyn pathology in brains of mice after unilateral injection of recombinant mouse αsyn fibrils into SN, Str, or EC. We confirmed the purity of recombinant αsyn monomer and fibrils used in this study; they didn’t contain any contaminants (Additional file 1: Figure S1). Using highly sensitive immunohistochemistry (IHC) with anti-phosphorylated αsyn (psyn) antibody 1175, we evaluated αsyn pathology in the brains at 1 month after injection. The distribution of αsyn pathology observed in these mice is illustrated in Figure 1. In mice injected into SN (Figure 1A), abnormal psyn pathology was restricted mainly to SN (3.08 mm posterior to bregma), amygdala (1.58 mm posterior to bregma), and stria terminalis (0.02 mm anterior to bregma) of the hemisphere on the injection side. In these mice, psyn was accumulated in neurites and soma (Figure 2A). In mice injected into Str, psyn pathology was widely distributed bilaterally throughout the brain, including Str (0.26 mm anterior to bregma), amygdala (1.58 mm posterior to bregma), SN (2.70 mm posterior to bregma) and cortex (Figure 1B). Psyn pathology was accumulated mainly in neurites, and partly in soma (Figure 2B). Injection of αsyn fibrils into EC induced severe psyn pathology in EC (3.52 mm posterior to bregma), dentate gyrus (3.52 mm posterior to bregma), hippocampal CA3 region (1.94 and 3.52 mm posterior to bregma), fimbria (1.94 mm posterior to bregma), and septal nuclei (0.02 mm anterior to bregma) on the injection side, as well as moderate psyn pathology in hippocampus on the contralateral side (Figure 1C). Psyn pathology was mainly observed in neurites and perinuclear regions (Figure 2C). No such psyn accumulation was detected in αsyn KO mice injected with αsyn fibrils into Str (Additional file 1: Figure S2A). Thus, there are major differences among these mice in the development and spread of αsyn pathologies, demonstrating that the propagation pattern depends upon the injection site.

To investigate whether other pathologies are also induced by the injection of αsyn fibrils, we performed IHC analysis using anti-tau, anti-TDP and anti-Aβ antibodies. No tau pathology was observed in mice injected into SN (Figure 3A). However, surprisingly, in mice injected into Str, pS396-positive dot-like structures were observed in Str, amygdala, and cortex (Figure 3B). Anti-phosphorylated tau (ptau) antibody AT8 also stained these structures in Str (Figure 3C). Similar ptau-positive dot-like structures were also observed with anti-pS396 antibody in EC, dentate gyrus and CA3 of the mice injected into EC, and were most frequent in CA3 (Figure 3D). Similar staining was observed in CA3 and dentate gyrus with AT8 antibody (Figure 3E). Furthermore, phosphorylated TDP-43 was also accumulated in mice injected into Str and EC (Figure 4B,C), although it was not detected in mice injected into SN (Figure 4A) at 1 month after injection. Aβ pathology was never observed in αsyn fibril-injected mice, regardless of injection site (Additional file 1: Figure S3). The tau and TDP-43 pathologies differed from psyn pathology in both shape and localization; most psyn pathologies were not colocalized with ptau-positive structures and the overlap was small (Figure 5).

To confirm the accumulation of these proteins and to analyze them biochemically, we next investigated sarkosyl-insoluble fractions of these mice brains at 3 months after injection into SN, Str or EC (Figure 6). Sarkosyl-insoluble psyn was detected in both the right and left hemispheres of all these mice, though it was more abundant on the injection side. The accumulation was most abundant on the injection side (right brain) in mice injected into Str, and less abundant on the uninjected side in mice injected into SN or EC. The banding patterns of sarkosyl-insoluble psyn were identical among these mice, regardless of the injection site, and were indistinguishable from that of DLB brain (Figure 6 upper). Anti-mouse αsyn antibody showed the same banding pattern as psyn antibody (Figure 6 middle). The 15, 22, 30 and 35 kDa bands correspond to monomer, monoubiquitinated αsyn, dimer and ubiquitinated dimer, respectively. Moreover, sarkosyl-insoluble ptau was detected in the right hemisphere of mice injected into Str, where the most abundant tau inclusions were observed (Figure 6 lower). On the other hand, αsyn and tau accumulations were not observed in αsyn KO mice injected with fibrils into Str (Additional file 1: Figure S4). These results indicate that inoculation of αsyn fibrils converted mouse αsyn at the injection sites to an abnormal form, that this change propagated from the injection site to the contralateral side of the brain, and that inoculation into Str also induced tau pathology.

Next, we analyzed motor and cognitive functions of these mice at 3 months after injection (Figure 7). Mice injected with αsyn fibrils into SN and Str showed poorer performance on the rotarod test compared with control mice injected with soluble αsyn (Figure 7A). Mice injected into SN also performed poorly on the wire hang test (Figure 7B). Cognitive dysfunction was not observed in any group in the Y-maze test (Figure 7C).

Finally, to examine whether insoluble αsyn induced in WT mice shows prion-like propagation behavior, we assessed the transmissibility of insoluble αsyn prepared from fibril-injected WT mouse brains. In brief, sarkosyl-insoluble αsyn was prepared from WT mouse brains injected with recombinant αsyn fibrils and was injected into Str of other WT mouse brains (Figure 8A, B). Induction and propagation of psyn pathology were examined by IHC. At 3 months after injection, psyn pathology was observed in Str (0.26 mm anterior to bregma) and had also propagated to amygdala (1.46 mm posterior to bregma) and SN (3.08 mm posterior to bregma) (Figure 8C). The distribution of psyn pathology is illustrated in Figure 8D. These data clearly showed that insoluble αsyn derived from WT mice injected with αsyn fibrils exhibits prion-like transmissibility.

Luk et al. and we have established that insoluble αsyn shows prion-like propagation behavior in WT mouse brain 20 , 21 , but the mechanism of spreading remains poorly understood. In this study, we investigated the spread and distribution pattern of psyn pathology in mouse brain injected with recombinant αsyn fibrils into three different brain regions: SN, Str, and EC. We assessed the distribution at 1 month post injection by using highly sensitive IHC. Pretreatment of brain sections with formic acid and heat enabled detection of psyn pathology at only 1 month after injection. When αsyn fibrils were injected into SN, psyn pathology only appeared in the central nucleus of amygdala and stria terminalis, which are located far from SN, while there was no detectable psyn pathology around SN (Figure 1A). Amygdala is connected with SN 30 , and stria terminalis serves as a major output pathway of the central nucleus of amygdala. These findings strongly indicate that spreading of psyn pathology does not occur by simple diffusion or nonspecific transport. In the case of injection into Str, psyn pathology was observed in amygdala, SN and a wide range of cortices (Figure 1B). Str has direct projection to SN and amygdala 31 , and many parts of the neocortex innervate the Str 32 . Injection into EC induced pathology in EC, dentate gyrus, hippocampal CA3, fimbria and septal nucleus (Figure 1C). Dentate gyrus receives projection from EC via the perforant pathway, and septal nucleus and fimbria have direct connections with hippocampus. Therefore, the data strongly suggest that propagation of pathological αsyn occurs via axonal transport and a trans-synaptic pathway, in accordance with reports that αsyn fibrils can be internalized by neurons and transferred from axons to second-order neurons in culture 33 and in animal models 34 , 35 . In patients with sporadic PD, the distribution of Lewy bodies and Lewy neurites seems to spread retrogradely 18 , 36 . In the present study, focusing on Str and SN, injection with αsyn fibrils into Str induced αsyn pathology in SN at 1 month after injection (Figure 1B). However, injection into SN did not induce pathology in Str at that time (Figure 1A), and pathology only became apparent in Str at 3 months after injection 21 , indicating there is a dominant direction of spread. Propagation from EC to dentate gyrus via the perforant pathway might occur via anterograde transport (Figure 1C). At least under our experimental conditions, propagation of αsyn seems to occur via both anterograde and retrograde transport processes. Thus, the predominant direction of spread presumably depends on cell types or brain areas. Similarly, tau is also reported to propagate via a trans-synaptic pathway in animal models 37 , 38 . Thus, axonal transport and trans-synaptic transport appear to be common pathways of propagation of intracellular aggregated proteins.

In addition, we found that tau and TDP-43 accumulation also occurred in WT mice injected with αsyn fibrils into Str and EC at 1 month after injection (Figures 3B-E and 4). The morphological patterns of tau and TDP-43 accumulation were apparently different from that of αsyn pathology (Figures 2, 3 and 4) and there was little colocalization (Figure 5), as in DLB brains 39 , 40 and a DLB mouse model 41 . Recently, Guo et al. reported that there are two strains of recombinant αsyn fibrils, strains A and B, and the two strains differently affect tau inclusion formation 42 . They reported that strain A (preformed fibrils) only infrequently induced tau inclusions and psyn pathology showed little colocalization with tau inclusions, whereas strain B (generated through repetitively seeded fibrillization in vitro) efficiently induced tau inclusions that were highly colocalized in neurons. The αsyn fibrils we used in this study are similar to strain A, and in agreement with their work 42 , we also detected small amounts of tau inclusions that showed little colocalization with αsyn pathology. In addition, tau accumulation was not observed in αsyn KO mice injected with αsyn fibrils by biochemical analysis (Additional file 1: Figure S4) or IHC (Additional file 1: Figure S2B). Thus, tau accumulation was induced by αsyn accumulation, and this occurred through a synergistic effect rather than a cross-seeding effect. The tau accumulation might be caused by a secondary effect of αsyn accumulation, such as dysfunction of cellular activity 43 , 44 or abnormality in protein degradation machinery 45 .

Biochemical analysis clearly showed that accumulated αsyn was phosphorylated and ubiquitinated similarly to that in DLB brain, regardless of injection site (Figure 6). This indicates that injection with the same fibrils as seeds induces αsyn aggregation in the same fashion, as is the case with prion strains.

Furthermore, αsyn fibril-injected mice showed modest motor abnormalities compared to the monomer-injected mice at 3 months after injection (Figure 7). This strongly suggests that propagation of psyn pathology induced motor phenotypes, although we could not detect cognitive dysfunction in the Y-maze test. It is possible that this is because we injected αsyn fibrils unilaterally, and the functions of the contralateral side of brain might be well maintained. In our previous study, we could not detect any abnormalities in fibril-injected mice at 6 months after injection 21 . The discrepancy may have arisen from differences in the test conditions, because in the previous study, we used female mice injected with human αsyn fibrils, whereas in this study we used male mice injected with mouse αsyn. Mouse αsyn fibrils propagate more efficiently in WT mice than do human fibrils 21 , and we think this was the main reason why we could detect motor abnormalities in the present study.

We next examined if insoluble αsyn accumulated in WT mice shows transmissibility. Our results demonstrate that insoluble αsyn accumulated in WT mice can induce αsyn pathology in other WT mice (Figure 8C, D), analogously to prion transmission. We also examined intraperitoneal injection or oral administration with αsyn fibrils into WT mice (see Materials and methods), but failed to detect any psyn pathology in the central nervous system at 6 months or 14 months after injection, respectively (data not shown).

Intracerebral injection with αsyn fibrils into WT mouse brains enables to induce phosphorylated αsyn pathology and the distribution of pathology depends on the injection sites. Furthermore, αsyn pathology has a synergistic effect on tau and TDP-43 aggregation. We conclude that αsyn fibrils have prion-like transmissibility and it might spread via axonal and trans-synaptic transports in mouse brains.

The authors declare that they have no competing interests.

MMS performed stereotaxic surgery, biochemical and IHC studies and wrote the manuscript. MK and AS performed IHC analysis and behavioral tests. TN and MH helped for interpretation of data. HA provided antibodies and helpful advice. MH performed study design and wrote the manuscript. All authors read and approved the final manuscript.