Lysates from symptomatic TgA53T mice and αS PFF induce α-synucleinopathy with similar end-stage distribution.
Although the intracortical/intrastriatal (IC/IS) transmission model shows the induction and spread of αS pathology in TgA53T mice6, studies show that injections of pathogenic αS lysates or preformed αS fibrils (PFF) in multiple areas, including the skeletal muscle, can induce widespread αS pathology over time9. Thus, we examined whether the inoculation site impacts the onset and spread of αS pathology in TgA53T mice. Because brain stem (BrSt) neurons are affected early by αS pathology in both human PD10 and in the TgA53T mouse model8, we compared the disease produced by the IC/IS inoculation with the BrSt injections of αS PFF. We injected young, disease-free 3–6 month-old TgA53T mice (all from line G2-3 unless noted) with end-stage lysate (ESL or S3000, a crude 3000xg pooled lysate) from BrSt/Spinal Cord (SpC) of end-stage (ES) affected TgA53T mice or lysates from disease-free (asymptomatic lysate, ASL) 5–6 months old TgA53T mice (Fig. 1a-c). Immunoblot analysis (Fig. 1b) show that both ASL and ESL contain similar amount of total αS but ESL contains greater amount of pS129αS, consistent with pathology in ES TgA53T mice. Like that seen by Luk et al. (2012), the mean of survival for bilateral IC/IS injected mice was 98 ± 8.6 days post-inoculation (dpi) (Mean ± SD) (Fig. 1d, Supplemental Fig. 1). Unilateral inoculation of ESL into the BrSt resulted in earlier onset of motor dysfunction with the mean survival of 71.3 ± 18.7 dpi (Fig. 1d, S1a), a significantly shorter disease time course than with IC/IS dual injections. Inoculation of αS PFF into BrSt leads to mean survival time (77.4 ± 13.16 dpi, n = 5) that is like the ESL inoculation (Fig. 1d, S1a). All the mice injected with ASL or saline remained disease free at ~ 150 dpi (Fig. 1d) except one ASL-inoculated animal who succumbed to disease prematurely died at ~ 9 months of age, a time point where normal aged TgA53T mice start to show disease8. Thus, it is likely that this animal developed disease independently of the injected lysate.
Biochemical analysis of the ESL inoculated animals confirms that the ESL, but not ASL, leads to αS pathology (Fig. 2a). Analysis of Triton X100-detergent soluble and insoluble fractions reveal the accumulation of insoluble high molecular weight (HMW) species of αS (Fig. 2a, asterisks) in the SpC of ESL injected mice. In the ESL injected mice, insoluble pS129αS accumulates in BrSt and SpC, corresponding to the regions most affected by αS pathology (Fig. 2a). Consistent with lack of αS pathology, pS129αS does not accumulate in ASL (Fig. 2a) or Saline (not shown) injected mice. This pattern of αS pathology is equivalent to that seen with the ES TgA53T mice that normally develop α-synucleinopathy from aging8.
Immunohistochemical detection of pS129αS demonstrates robust staining in multiple brain tissues from IC/IS- and BrSt-injected end-stage animals (Fig. 2b). Consistent with the lack of motor phenotype, no pathological alterations are noted in ASL-inoculated mice (Fig. 2b) or saline-injected animals (data not shown). Significantly, despite the differences in the site of inoculation (IC/IS vs BrSt), the overall patterns of pathology in the ES mice were comparable to each other with most abundant pathology in the BrSt and SpC (Fig. 2b), a pattern very similar to what is seen in the TgA53T mice that naturally develop αS pathology with aging8. However, we noted that IC/IS injected animals exhibit more pS129αS pathology in CTX compared to BrSt injected animals (Fig. S1b). To determine if components of lysates other than pathogenic αS could be responsible for the distribution of the pathology, we also examined animals that were BrSt injected with αS PFF. Our results show that the pattern of pathology seen in the ESL inoculated animals is very similar to animals injected with αS PFF into BrSt (Fig. 2b, S1b). Moreover, the pattern of αS pathology induced by BrSt injections are also very similar to the pathology achieved in this line of mice by peripheral, intramuscular (IM) inoculation of αS PFF9,11. To determine if the prominent subcortical pathology is unique to TgA53T(G2-3) line or occurs in another line of TgA53T, we performed IC/IS injections on TgA53T(H5), which express 50% less αS than TgA53T(G2-3)8,12. Because of lower transgene expression, injected TgA53T(H5) animals developed progressive motor phenotype later (~ 150 dpi) than the TgA53T(G2-3) (~ 100 dpi) following IC/IS injections of αS PFF. Neuropathological analysis shows that, as with TgA53T(G2-3), αS pathology at ES mice was most abundant in BrSt/SpC region (Fig. S2). Significantly, the TgA53T(H5) animals also showed a more widespread distribution of αS pathology, including αS pathology in the hippocampus (Fig. S2a). In summary, in all animals inoculated with ESL or αS PFF showed most severe BrSt and SpC αS pathology, regardless of initial sites of injections.
Collectively, these results indicate that in TgA53T mice, the ES distribution of αS pathology is independent of inoculation site and likely determined by factors other than local seeding by pathogenic αS. Moreover, because our Tg models show highest levels of transgene expression in the CTX8,11,13, relative lack of forebrain pathology is not because of insufficient transgene expression in this area. While previous studies with the IC/IS injection model show that cortical and striatal αS pathology occurs prior to BrSt or SpC6, we observe more prominent early subcortical pathology with the BrSt inoculation. Thus, following the BrSt inoculation, the appearance of pS129αS over time shows that initial αS pathology develops proximal to the inoculation site where αS pathology is seen in the pons by 30 dpi (Fig. 2c). By 45 dpi, regions rostral and caudal to the BrSt inoculation sites exhibits αS pathology (Fig. 2c). Thus, we propose that following BrSt injection of pathogenic αS, the pathology spreads both rostral and caudal directions but most prominent pathology occurs in subcortical regions (Fig. 2d). Analysis of IC/IS injected TgA53T(H5) animals at intermediate stage (90 dpi) shows the presence of pS129αS in the CTX but more obvious pS129αS pathology is seen in the BrSt and SpC areas (Fig. S2b). Overall, current results collectively show that regardless of the initial injection site, subcortical areas (e.g. BrSt, SpC) exhibit earliest αS pathology and culminates in most severe αS pathology at ES TgA53T mice.
Neuroinflammation follows onset of α-synucleinopathy.
It has been proposed that inoculation with pathogenic αS could lead to early neuroinflammation that may promote further propagation of α-synucleinopathy14. Thus, we examined the spatial-temporal relationship between αS pathology and neuroinflammation. First, we confirmed that at ES, animals following BrSt injection exhibit coincident αS pathology and neuroinflammation. Tissue sections from BrSt-injected animals were stained for microglia (Iba1; Fig. 3a) and astrocytes (GFAP; Fig. 3b). Based on the morphology of the cells, no obvious neuroinflammatory changes are seen in absence of significant αS pathology, such as in CTX (Fig. 3a,b) and in animals injected with saline (not show) or ASL (Fig. 3a,b). However, in ES animals following ESL injection, areas with significant αS pathology (BrSt, SpC) exhibit abundant microglia and astrocytes with highly activated morphology. Immunoblotting for Iba1 and GFAP reveals no change in Ctx, but a significant increase in BrSt and SpC of ESL-injected mice at ES disease (Fig. 3c,d; Fig. S3).
To better determine the spatial-temporal relationship between the onset of microglial activation and αS pathology, the sections from BrSt inoculated mice harvested at 15-, 30-, and 45-dpi that were evaluated for αS pathology (Fig, 2c) were stained for microglia (Iba-1) and astrocytes (GFAP) (Fig. S4). With BrSt inoculation, αS pathology is seen in the pons by 30 dpi (Fig. 2c). However, there is no obvious signs of astrocytic or microglial activation at 30- or 45-dpi (Fig. S4). Because αS pathology is still sparse at 45 dpi, we performed double immunofluorescence analysis of pS129αS with Iba1 or GFAP (Fig. 4) to determine if presence of αS pathology could be associated with local changes in inflammatory changes. Double immunofluorescence analyses show that that while αS pathology in ES mice are clearly associated with nearby reactive astrocytes and microglia, such glial responses do not accompany αS pathology at 30- or 45-dpi (Fig. 4b,c). Similar analysis of early-stage mice following IM αS PFF injections show that initial onset of pS129αS pathology in SpC, occurring as early as 15–30 dpi, is not accompanied by obvious glial activation (Fig. S5). While other A53T Tg models show αS pathology in astrocytes15,16, we do not observe significant pS129αS in astrocytes at ES mice (Fig. 4)17.
Collectively, our results indicate that glial activation, particularly activation of microglia, is only evident after substantial αS pathology is established. Similar pattern of initial αS pathology followed by neuroinflammation was reported with IM inoculation of M83 line18. Thus, while microglia and astrocytes may modulate neuronal spreading of αS pathology or neuronal survival, our results indicate that it is unlikely that the neuroinflammation is a significant factor in the initial onset of αS pathology in TgA53T model of α-synucleinopathy.
Both highly soluble and insoluble fractions induce αS pathology
Because αS fibrils and disease associated aggregates in TgA53T mice are detergent insoluble9,13, we tested whether the induction of αS pathology by the ESL is mediated by the insoluble αS species. The pathogenic S3000 ESL was centrifuged at 150,000xg to obtain highly soluble (supernatant, S150) and insoluble (pellet, P150) fractions (Fig. 5a). Biochemical analyses of the fractions show that very little αS is found in P150 from asymptomatic animals (Fig. 5b). Moreover, most of the pS129αS, representing the overt αS pathology, is highly enriched in the P150 fractions from the symptomatic mice (Fig. 5b). In contrast, the amount of total αS is the same in the S150 fractions, regardless of the disease state of the animal (Fig. 5b). Despite the high levels of total αS, the levels of pS129αS in S150 is lower than in the P150 with no differences between the disease state of the animal (Fig. 5b). Thus, we expected that αS pathology would be selectively induced by the P150-ESL fraction.
Following BrSt injections of P150 and S150 fractions, we were surprised to find that, regardless of the solubility of the inoculated material, inoculated TgA53T mice developed motor dysfunction leading to premature death with average lifespans of 75 ± 7 dpi for P150 fraction and 88 ± 3 dpi for S150 fraction (Fig. 5c). While 2 out of 7 subjects injected with S150 did not develop disease phenotype by 150 dpi, for those animals that died prematurely, there was no differences in the average lifespan (Fig. 5c, p = 0.1397) between S150 and P150 injected groups.
Histological (Fig. 5d) and biochemical (Fig. S6) analyses of ES P150- and S150-injected animals demonstrate virtually indistinguishable αS pathology both in the spatial pattern and severity of αS aggregation (compare Figs. 2 and 5, S6). Aggregation of αS occurred throughout the CNS but was most robust in the BrSt and SpC (Fig. 5d). Biochemical analysis of brain tissue also reveals accumulation detergent insoluble αS species in the insoluble fraction from BrSt and SpC (Fig. S6a) and dramatic increases in pS129αS in both BrSt and SpC (Fig. S6b). Finally, based on immunostaining for Iba1 and GFAP, the distribution and intensity of neuroinflammation is similar in S150- and P150-BrSt-injected animals (Fig. 5d), and is nearly identical to that observed in ESL-injected mice (see Fig. 3). Thus, by three broad metrics: survival, IHC, and biochemistry, the α-synucleinopathy and disease induced by injection of all three end-stage lysates (S3000, S150, P150) are essentially identical.
Since S3000 ASL does not induce disease and there is no obvious differences between S150 from ASL and ESL on our immunoblot analysis, S150 from ESL may contain soluble pathogenic αS conformers that is SDS labile. Thus, incomplete penetrance of end-stage S150 to cause disease could be due to the possibility that toxic αS assemblies in S150 fractions are more prone to degradation following cellular uptake compared to more mature insoluble toxic αS species in P150 (Fig. 5b). As an initial test of this hypothesis, we performed dot blot analysis of P150 and S150 for αS oligomers, using FILA119,20 and OC antibody21 (Fig. S7a,b). We previously showed that FILA1 can recognize both soluble and insoluble αS aggregates derived of TgA53T model20 and OC antibody selectively recognizes mature fibrils21. Our results show that while levels of FILA1 + oligomers are similar between P150 and S150, the levels of OC + oligomers are significantly more abundant in P150 (Fig. S7a,b). We also subjected S150 and P150 to proteinase K treatment as more compact mature aggregates should be more resistant to proteinase K treatment. The results show that αS in P150 is more resistant to proteinase K proteolysis than αS in S150 (Fig. S7c). Thus, we conclude that the presence of FILA1 + oligomers/aggregates in both S150 and P150 are pathogenic. Further, more labile FILA1 + oligomers in S150 may be responsible for the partial penetrance of S150 in inducing disease.
Microsomes from symptomatic TgA53T mice induce rapid α-synucleinopathy.
We previously showed that α-synucleinopathy in TgA53T mice is associated with accumulation of αS oligomers/aggregates in endoplasmic reticulum (ER) and chronic ER stress contributes to neurodegeneration13,20. Analyses of TgA53T mice inoculated with S3000 ESL show that αS pathology in these mice are also associated with signs of chronic ER stress (Fig. S8), suggesting that ER stress occurs even when α-synucleinopathy is induced by exogenous inoculation of pathogenic αS. We recently showed that ER enriched microsomes from TgA53T mice, containing αS aggregates, are highly toxic to cultured neurons and aggressively induces αS aggregates and cell death in cultured neurons22. Thus, we examined whether the microsome fraction from symptomatic TgA53T mice can induce pathology following BrSt injection.
Microsomes from pooled BrSt/SpC or CTX from the symptomatic TgA53T mice were used for BrSt injections of 3–4 mos old TgA53T mice. Analysis of the fractions show that pS129αS is enriched in microsomes (P100 or ER) from BrSt/SpC where very little pS129αS is seen in P100 from CTX (Fig. 6a). Analysis of organelle markers show that P100 fraction is also enriched in the ER marker (Grp78/BiP). Analysis of injected animals show that P100 from symptomatic TgA53T mice develop progressive motor abnormalities by ~ 50 dpi while the animals inoculated with P100 from CTX did not show any disease phenotype (Fig. 6b). Significantly, microsome-inoculation leads to onset of motor symptoms much faster than any of the other fractions tested in this study (Compare Figs. 1d,5c & 6b), indicating that microsome associated αS oligomer/aggregates are highly pathogenic in vivo. Neuropathological analysis show that the early motor deficits were indeed associated with significant αS pathology (Fig. 6c) with the pattern that is seen with other αS fractions. Collectively, these results provide in vivo confirmation of highly pathogenic nature of ER-associated αS oligomer/aggregates20,22.