Pathological α-Synuclein Transmission Initiates Parkinson-like Neurodegeneration in Non-transgenic Mice

Parkinson’s disease (PD) is a multisystem neurodegenerative disorder characterized by two major disease processes: the accumulation of intraneuronal Lewy-bodies/Lewy-neurites (LBs/LNs) containing misfolded fibrillar α-Syn, and the selective degeneration of midbrain dopamine (DA) neurons in the substantia nigra pars compacta (SNpc) leading to bradykinesia, tremor, and postural instability (1). The etiology of these processes remains unclear, although in familial PD, autosomal dominant α-Syn gene mutations or amplifications directly link α-Syn dysfunction to disease causation (2, 3). Moreover, despite the observation that α-Syn pathology is present in virtually all sporadic and familial PD patients (4) and that its distribution correlates with clinical symptoms (5–7), the precise relationship between α-Syn misfolding and DA neuron loss remains controversial because neither transgenic nor neurotoxin-based animal models of PD fully recapitulate the DA neuron degeneration, motor deficits, and the LB/LN-like α-Syn pathology seen in PD (8, 9).

Converging lines of evidence indicate that misfolded fibrillar forms of neurodegenerative disease-related proteins self-propagate and spread between interconnected central nervous system (CNS) regions, suggesting that cell-to-cell transmission of pathological proteins contributes to disease progression (10–15). Indeed, the pattern of LB/LN accumulation in human PD is highly stereotypical, consistent with propagation of disease over functional networks, suggesting that the transmission of pathologic α-Syn and its corruption of endogenous protein plays a central role in PD pathogenesis and progression (16, 17). Moreover, reports that embryonic mesencephalic neuron grafts in PD patients develop LBs many years after grafting (18, 19) further support the transmissibility of pathologic α-Syn. Consistent with this hypothesis, intracerebral inoculation of α-Syn pre-formed fibrils (PFFs) comprised of recombinant protein accelerates the onset of neurological symptoms and death in transgenic mice overexpressing human A53T mutant α-Syn by enhancing the conversion of endogenous α-Syn into pathological LBs/LNs (20). While these data implicate α-Syn misfolding and fibrillization as triggering events upstream of neuronal dysfunction/death, minimal DA cell loss was observed, possibly reflecting alterations in the pattern of α-Syn transgene expression and the rapid demise of the inoculated mice.

Because synthetic α-Syn PFFs efficiently seed the aggregation and fibrillization of soluble endogenous α-Syn in primary neuronal cultures generated from wildtype (wt) mice (21), we asked if transmission of LB/LN pathology might occur in wt mice in vivo. To answer this, we first determined whether PFFs assembled from recombinant mouse α-Syn initiate pathological α-Syn conversion when introduced into young wt (C57BL6/C3H) mice by stereotaxic injections targeting the dorsal striatum (Fig. S1), a region receptive to PFF uptake (20) and interconnected with multiple CNS nuclei, including midbrain DA neurons. Deposits of hyperphosphorylated α-Syn (pSyn), a marker of human LBs/LNs (22), were visible at the injection site within 30 days of a single unilateral PFF inoculation (Fig. 1A,B,G). Intraneuronal α-Syn accumulations, primarily diffuse LN- and LB-like inclusions, were also present in several areas directly interconnected to the striatum that accumulate LBs in human synucleinopathies, most prominently in cortical layers IV-V, and olfactory bulb (Fig. 1C,D), confirming the pathological conversion of endogenous mouse α-Syn in wt mice. At 30 days post-injection (dpi), pSyn positive LB-like accumulations were exclusively ipsilateral to the injection site with the exception of the amygdala, to which the striatum projects bilaterally (23), suggesting that cell-to-cell transmission followed inter-neuronal connectivity.

LB/LN-pathology within affected regions showed markedly increased pSyn immunoreactivity in mice examined 90 and 180 dpi (Fig. 1E,F,G). Additional populations, such as the contralateral neocortex, also developed LBs/LNs indicative of progressive spread to CNS regions (Fig. 1F–I, Table S1). Mapping of pSyn pathology in mice at 30, 90, and 180 dpi (Fig. 1G) revealed a time-dependent dissemination of LBs/LNs between 30–180 dpi with sequential involvement of populations initially unaffected at 30 dpi including ventral striatum, thalamus, and occipital cortex, along with commissural and brainstem fibers. Although inclusions were absent from multiple regions including hippocampus, septum, and cerebellum (Fig. 1G and Table S1), stereotaxic inoculation of equal quantities of PFFs into the hippocampus resulted in massive accumulation of α-Syn pathology (Fig.S2), indicating that neurons in this region also are susceptible to transmitted pathologic α-Syn. Thus, propagation of LBs/LNs is connectivity-dependent and provides an explanation for the non-uniform distribution of LBs/LNs observed in α-Syn PFF-seeded animals. In contrast, monomeric α-Syn or PBS injections did not result in α-Syn pathology at any time point examined (Fig. 1J and Fig.S2).

Striatal PFF inoculation led to abundant α-Syn pathology in tyrosine hydroxylase (TH) positive SNpc DA neurons (Fig. 1K,L) that project to medium spiny neurons of the dorsal striatum and whose selective degeneration is a major contributor to motor deficits in PD (24). Furthermore, α-Syn pathology in PFF-inoculated mice colocalized with key markers of LBs, including ubiquitin, heat-shock protein 90, and the amyloid-binding dye thioflavin S, indicating that they share common properties with authentic LBs/LNs in human PD (Fig.S3).

SNpc α-Syn pathology developed progressively following PFF injection, evolving from pale cytoplasmic accumulations at 30 dpi to dense perinuclear LB-like inclusions by 90 and 180 dpi, particularly among ventromedial SNpc populations (Fig. 2A–C). Because the nigrostriatal pathway is exclusively unilateral, LB/LN pathology in the SNpc was confined to the injected hemisphere at all time points examined (Figs.1G and 2A–D). In addition to being absent in ventral tegemental (VTA) neurons, α-Syn pathology was not observed in DA-expressing olfactory glomerular interneurons and arcuate cells of the hypothalamus (Fig.S4). Similarly, norepinephrine (locus coeruleus) and serotonergic (Raphe nucleus) populations also remained unaffected at 180 dpi, suggesting that monoaminergic neurons are not equally susceptible to pathological α-Syn spread.

α-Synuclein pathology in the SNpc was accompanied by the gradual and unilateral loss of TH-immunoreactivity and neurons (Fig. 2F–I), suggesting that intraneuronal α-Syn inclusions lead to DA neuron loss. 27.7% of SNpc DA neurons colocalized with pSyn pathology at 30 dpi (Fig. 2K), a stage at which SNpc and VTA DA neuron number were similar in both hemispheres (Fig. 2L,M). SNpc but not VTA DA neurons developed LBs/LNs, confirming transmission of misfolded α-Syn via network projections and suggesting that connectivity is a determinant of susceptibility. The proportion of inclusion-bearing SNpc DA neurons did not increase further, but instead declined over time. There was a concomitant decrease of SNpc DA neurons by 15% and 35% at 90 and 180 dpi, respectively, suggesting that LBs/LNs form prior to SNpc DA neuron loss. Moreover, DA neurons in the contralateral SNpc and the VTA of both hemispheres (where inclusions were present in <0.2% of neurons; Fig. 2K) were indistinguishable from those in PBS- or monomer-injected controls, non-injected age-matched mice, and PFF-injected Snca^−/− mice, all of which did not develop α-Syn deposits (Fig. 2D,K–M). Thus, accumulation of pathologic α-Syn appears to be upstream of and directly linked to SNpc DA neuron loss.

Concomitant with the progressive loss of SNpc neurons, striatal concentrations of DA and its metabolites in the injected hemisphere showed progressive reductions following PFF-inoculations, whereas contralateral striatal DA levels remain unaltered, and noradrenaline (NA) levels were unchanged bilaterally (Figs.3A,B and S5A). Immunoblot analyses confirmed reductions in TH and dopamine transporter (DAT) expression in the ipsilateral striatum (Fig. 3C) consistent with the loss of nigrostriatal DA innervation. Despite the loss of DA terminals, D1 and D2 dopamine receptor levels remained constant (Fig. 3C, and S5C), suggesting the preservation of striatal medium spiny projection neurons.

Given the PD-like Lewy pathology, SNpc neuron loss, and reduced striatal DA levels, PFF-inoculated wt mice might be expected to exhibit altered motor function, as in PD. Although no gross motor or behavioral abnormalities were observed up to 180 dpi, rotarod testing revealed significant and progressive performance deterioration in PFF-injected animals, indicative of impaired motor co-ordination and balance (Fig. 3D, left). Furthermore, mice with α-Syn pathology performed poorly on the wire hang test, a complementary measure of motor strength and co-ordination, declining 53% from baseline at 30 dpi and 81% at 180 dpi (Fig. 3D, middle). However, open field activity remained unchanged, indicating overall motor activity was not significantly altered (Fig. 3D, right).

As previously noted, intrastriatal PFF-inoculated mice did not develop hippocampal pathology (Fig. 1G), and no changes in memory function were observed as assessed using the Y-maze (Fig. 3E). This corroborates our interpretation that declining neuronal function is correlated with increasing α-Syn pathology in affected neurons. Moreover, in keeping with the relative sparing of mesolimbic projections, no significant differences were found between PFF-treated and control mice in the tail suspension test (Fig. 3E, right) a measure of depression-like behavior. The correlation between α-Syn pathology and neuronal deficits was also seen in striatal PFF-inoculated Snca^+/− mice, where at 180 dpi both SNpc α-Syn pathology and motor impairments were attenuated relative to wt animals (Figs.3D,E and S6A). Finally, striatal inoculation into wt CD1 (Fig.S6B) andC57BL6/SJL mice (Table S1) also resulted in transmission of α-Syn pathology, suggesting that cell-to-cell spread of LBs/LNs is independent of mouse genetic background.

In summary, we demonstrate that a single intrastriatal injection of synthetic misfolded α-Syn seeds into wt mice initiates a neurodegenerative cascade characterized by the accumulation of intracellular LB/LN pathology, selective loss of SNpc DA neurons, and impaired motor coordination. Thus, α-Syn pathology is sufficient to induce the cardinal behavioral and pathological features of sporadic PD.