Lewy bodies and neurites are the pathological hallmark of Parkinson's disease. These structures are composed of fibrillized and ubiquitinated alpha-synuclein suggesting that impaired protein clearance is an important event in aggregate formation. The A30P mutation is known for its fast oligomerization, but slow fibrillization rate. Despite its toxicity to neurons, mechanisms involved in either clearance or conversion of A30P alpha-synuclein from its soluble state into insoluble fibrils and their effects in vivo are poorly understood. Synphilin-1 is present in Lewy bodies, interacting with alpha-synuclein in vivo and in vitro and promotes its sequestration into aggresomes, which are thought to act as cytoprotective agents facilitating protein degradation. We therefore crossed animals overexpressing A30P alpha-synuclein with synphilin-1 transgenic mice to analyze its impact on aggregation, protein clearance and phenotype progression. We observed that co-expression of synphilin-1 mildly delayed the motor phenotype caused by A30P alpha-synuclein. Additionally, the presence of N- and C-terminal truncated alpha-synuclein species and fibrils were strongly reduced in double-transgenic mice when compared with single-transgenic A30P mice. Insolubility of mutant A30P and formation of aggresomes was still detectable in aged double-transgenic mice, paralleled by an increase of ubiquitinated proteins and high autophagic activity. Hence, this study supports the notion that co-expression of synphilin-1 promotes formation of autophagic-susceptible aggresomes and consecutively the degradation of human A30P alpha-synuclein. Notably, although synphilin-1 overexpression significantly reduced formation of fibrils and astrogliosis in aged animals, a similar phenotype is present in single- and double-transgenic mice suggesting additional neurotoxic processes in disease progression.
Parkinson's disease (PD) is a progressive neurodegenerative disorder of the central nervous system. At the histological level, PD is characterized by the presence of intraneuronal cytoplasmic inclusions described as Lewy bodies (1,2). In most cases, sensorimotor symptoms of PD are directly related to the distribution of Lewy bodies spreading from the lower brainstem and olfactory bulb into the substantia nigra, the limbic system and to the neocortex (3). Lewy bodies are mainly composed of alpha-synuclein (α-syn), a small 140 amino-acid protein abundant in various regions of the brain and enriched at presynaptic terminals (4,5). The importance of α-syn in PD is highlighted by rare autosomal dominant forms of PD associated with mutations in the α-syn protein (A30P, E46K, H50Q, G51D, A53T) or by its genomic multiplications (6–12). Increasing evidence suggests that α-syn accumulation has a detrimental effect on aggregate progression, which is linked to the formation of a complex equilibrium of its oligomeric, protofibrillar and fibrillar forms (13–16). Several intracellular mechanisms may be involved in α-syn fibrillization, namely: (i) α-syn mutations (17), (ii) post-translational modifications as C-terminal truncation (18,19), (iii) site-specific ubiquitination (20–22) and (iiii) dysfunctional protein degradation systems (23–25). Whether α-syn aggregate formation is caused by a dysfunctional protein clearance either of soluble, misfolded and/or post-translationally modified structures is still unknown.
In mammals, the ubiquitin–proteasome system (UPS) (26) and the autophagy–lysosome pathway (ALP) are the two major mechanisms involved in the degradation of α-syn (27). Whereas the UPS is the main pathway that mediates the degradation of α-syn under physiological conditions, autophagy is activated when the level of α-syn is pathologically increased (28).
In vitro studies showed that overexpression synphilin-1 (sph1), which co-localizes with α-syn in neurons and Lewy bodies (29,30), inhibits proteasomal degradation of α-syn (31), increases its aggregation and promotes formation of aggresomes (32–34), displaying different susceptibility to autophagic clearance pathways (35,36). Despite congruent findings in cellular models, results gained from overexpression of sph1 in transgenic α-syn animals are less consistent. Sph1 overexpression in A53T human α-syn transgenic mice reduced astrogliosis and neuronal degeneration, resulting in an increased life span (37). However, sph1 expression into the substantia nigra of A30P α-syn transgenic mice via viral-mediated gene delivery resulted in the loss of tyrosine hydroxylase positive neurons without misfolding of human A30P α-syn into proteinase K (PK)-resistant or thioflavin S positive structures (38). Although this acute model is valuable to analyze the local effect of sph1, the age-dependency of aggregation processes in brainstem regions displaying strongest pathology in the A30P α-syn transgenic mice (39) is not addressed.
In order to answer long-term effects of sph1, we crossed human wild-type sph1 transgenic mice (40) with the well-characterized human A30P transgenic mouse line (41–45). We analyzed the molecular relevance of sph1 in the development of the A30P-induced synucleinopathy.
In the present study, we report that overexpression of sph1 in A30P mice mildly delayed the onset of motor symptoms, which was paralleled by a strong reduction of markers for α-syn aggregation as its conversion into thioflavin S positive and PK-resistant mature fibrils. Additionally, the amount of soluble monomeric and fragmented α-syn was decreased while markers for aggresomes and autophagy were elevated, indicative of accelerated protein degradation. The lack of fibril pathology yet persistence of motor dysfunction in aged mice, supports the idea that fibrillized α-syn deposits are not solely linked to the phenotype progression.
Overexpression of synphilin-1 diminishes the accumulation of full-length and truncated soluble alpha-synuclein in A30P transgenic mice
Mice overexpressing the human A30P mutant alpha-synuclein (α-syn) under the Thy-1 promoter (A30P) (42) were previously characterized, showing an age-dependent loss of motor function and accumulation of α-syn (39,43,44). Mice overexpressing the human synphilin-1 (sph1) under the PrP promoter (sph1) were generated and described by our laboratory, displaying high transgenic sph1 level in brainstem neurons but without α-syn aggregation pathology (40). Here, we generated heterozygous double-transgenic mice (double), overexpressing both, human A30P α-syn and human sph1, by crossbreeding homozygous A30P, sph1, double-transgenic and non-transgenic (nTg) animals with C57BL/6 nTg littermates (Fig. 1A). Genotypes were identified using genomic DNA PCR and heterozygosity was confirmed via real-time quantitative PCR as previously described (40,42). We performed western blot analysis of brainstem tissues to evaluate the expression level of transgenic α-syn proteins (Fig. 1B). Consistent with previous observations (42), we detected an age-dependent increase of soluble monomeric human (39.0 ± 13.8%) and total α-syn (35.5 ± 16.0%) signals at 14 kDa in brainstem of single-transgenic A30P animals (P < 0.05, Fig. 1B and C). In contrast, there was no significant accumulation of human mutant A30P (4.7 ± 14.2%) or total α-syn (10.7 ± 12.7%) with aging in double-transgenic animals (P > 0.05, one way ANOVA, post-hoc LSD, Fig. 1B and C). Additionally, α-syn antibodies raised against the C-terminus and the NAC domain of α-syn are able to detect truncated α-syn species (46,47). Using these specific antibodies, we observed a strong decrease of truncated α-syn fragments in 18-months-old double-transgenic mice when compared with age-matched single-transgenic A30P mice (C-terminal: 69.7 ± 8.9%; NAC domain 12-kDa fragment: 91.8 ± 9.3% and 10-kDa fragment: 76.0 ± 9.8%; P < 0.01, Fig. 1D and E). We also observed a decrease of human (34.6 ± 6.6%; P < 0.01, Fig. 1D) and total (37.2 ± 6.1%) soluble full-length α-syn in 18-months-old double-transgenic mice (P < 0.05, Fig. 1E), thus suggesting differences in transgenic protein degradation in double-transgenic mice.
Synphilin-1 and alpha-synuclein are co-localized in aggresome-like structures in double-transgenic mice
We next analyzed the co-localization of human α-syn with sph1 in the brainstem of 18-months-old animals, a region displaying strongest aggregation pathology in single-transgenic A30P α-syn mice (39) (Fig. 2). Using double immunofluorescence approach with antibodies specific for human α-syn and for myc-tag of sph1, we detected neuronal labeling of human α-syn (Fig. 2C and D) and myc-tagged sph1 (Fig. 2B and D), co-localizing in double-transgenic mice (merged in Fig. 2D). Human A30P α-syn and sph1 were localized throughout the cytosol of neuronal cell bodies as previously described (39,40). We noticed a more diffuse cytoplasmic staining and small, punctate fluorescence α-syn signals in the neuropil of old A30P mice (Fig. 2C) when compared with age-matched double-transgenic animals (Fig. 2D). Strikingly, we found several large and round-shaped inclusions immunopositive for both α-syn and sph1 selectively in double-transgenic animals, resembling aggresome formations (Fig. 2D, merged panel). Co-localization of human α-syn with γ-tubulin (Fig. 2E; merged panel) validated its accumulation into aggresomes in double-transgenic mice according to previously established structural criteria (reviewed in 48,49).
Synphilin-1 overexpression decreases the amount of insoluble and fibrillized alpha-synuclein in A30P transgenic mice
Overexpression of human A30P α-syn is leading to an age-dependent increase of detergent-insoluble and fibrillized α-syn in transgenic mice (42,43). In order to separate and quantify soluble cytoplasmic proteins from SDS-insoluble aggregates, we performed sequential protein extraction using buffers with increased detergent strength as previously established for PD (50). SDS-insoluble aggregates were then solubilized in 2 M urea and α-syn levels were determined using SDS-gel electrophoresis. In the brainstem of 18-months-old animals, we observed an increase of insoluble α-syn in single-transgenic A30P and in double-transgenic animals when compared with nTg or sph1 single-transgenic controls (nTg: 4.3 ± 0.8%, sph1: 4.3 ± 2.2%, A30P: 100 ± 51.5%, double: 65.7 ± 22.1%; P < 0.05, one way ANOVA, post-hoc LSD, Fig. 3A). Although levels of insoluble α-syn were reduced in aged double-transgenic animals when compared with single-transgenic A30P mice, it did not reach significance (P > 0.05, one way ANOVA, post-hoc LSD, graph in Fig. 3A). In order to analyze whether insolubility is also linked to an increase in mature fibril formation at the ultrastructural level, we isolated α-syn filaments by sucrose gradient centrifugation and quantified immunogold labeled fibrils as previously described (51) using an antibody specific for both, human and rodent α-syn. As expected, the number of filaments per field (∼10 µm2) was significantly higher in single-transgenic A30P when compared with nTg mice (nTg: 2.32 ± 0.7, sph1: 0.2 ± 0.1, A30P: 8.1 ± 0.5, double: 0.8 ± 0.2; P < 0.001, one way ANOVA, post-hoc LSD, Fig. 3B). Importantly, overexpression of sph1 strictly reduced the level of α-syn fibrils (P < 0.01, one way ANOVA, post-hoc LSD, Fig. 3B) in double-transgenic mice. Relatively long α-syn filaments, immunodecorated with five or more gold particles were found in single-transgenic A30P mice only but not in double-transgenic mice (data not shown). Taken together, this implies that although α-syn is converted into insoluble structures, co-expression of sph1 profoundly reduces fibril formation in vivo.
Synphilin-1 overexpression reduces the amount of proteinase K-resistant alpha-synuclein, thioflavin S positive fibers and gliosis in A30P transgenic mice
Next, we examined characteristics of α-syn aggregates in the rostral part of the reticular formation, the most affected brainstem region in A30P animals (39,43). For this, sections of 18-months-old mice were subjected to proteinase K digestion and subsequently stained against human α-syn; untreated sections served as controls (Fig. 4 A and B). Using this approach, we observed a strong increase of proteinase K resistant aggregates in single-transgenic A30P mice (A30P: 3.9 ± 0.4, double: 0.6 ± 0.3; P < 0.001, Fig. 4A and B). To validate aggregation pathology, we stained adjacent sections with thioflavin S (Fig. 4C). We observed a strong increase of thioflavin S positive structures in single-transgenic A30P mice only (nTg: 0.7 ± 0.2, sph1: 0.4 ± 0.04, A30P: 4.6 ± 1.1, double: 0.6 ± 0.7; P < 0.001, one way ANOVA, post-hoc LSD, Fig. 4C). These sections showed a strong accumulation both of PK-resistant and thioflavin S positive profiles in dystrophic axons and neuronal soma (Fig. 4B and C). Consistent with the ultrastructural findings, the number of cells harboring proteinase K resistant α-syn, as well as thioflavin S positivity, were strictly reduced in double-transgenic animals (P < 0.01, one way ANOVA, post-hoc LSD, Fig. 4A–C). Reactive astrogliosis is a neuropathological change commonly observed in α-syn transgenic animals displaying aggregation pathology (52,53). Using immunohistochemical staining of astroglial specific glial fibrillary acidic protein (GFAP), we detected an increase in astrogliosis in aged single-transgenic A30P mice and in double-transgenic animals (nTg: 1.2 ± 0.7, sph1: 1.7 ± 0.9, A30P: 16.3 ± 0.3, double: 10.0 ± 0.6; P < 0.01, Fig. 4D). Paired comparison with age-matched single-transgenic A30P animals revealed a significant decrease of GFAP-positive astrocytes in double-transgenic mice (P < 0.01, one way ANOVA, post-hoc LSD, Fig. 4D). These observations clearly demonstrate a reduction of distinct α-syn aggregate markers and astrogliosis in double-transgenic mice, suggesting a modifying role of sph1 in clearance of misfolded α-syn.
Synphilin-1 overexpression increases the amount of ubiquitin-containing proteins in A30P transgenic mice
Abnormal accumulation of ubiquitin was previously reported in neurons presenting α-syn aggregation in PD (54,55) and A30P transgenic animals (39,42,43,52). We observed an accumulation of multiple small ubiquitin-positive aggregates in cell bodies and neuropil of single-transgenic A30P animals (Fig. 5A; arrows in A30P panel). In double-transgenic animals, however, we detected formation of few and relatively large ubiquitin-positive structures at the perinuclear region, resembling aggresomes (Fig. 5A; arrow in ‘double’). In order to quantify and characterize these inclusions, we analyzed high molecular weight ubiquitinated proteins in the brainstem of 18-months-old mice by using AGERA. This method allows the separation of proteins larger than 200 kDa using 1.5% agarose gel electrophoresis (56). We confirmed an increase of ubiquitin signals in double-transgenic mice (Fig. 5B). Similarly, immunoblots gained from high-percentage gels supported this notion as an accumulation of high-molecular weight ubiquitinated protein signals was detected in form of sequestered material at the top of the stacking gel (data not shown). We then investigated the amount of lower molecular weight ubiquitinated proteins in brainstem of young and aged mice using conventional denaturating SDS-gel-electrophoresis and immunoblot techniques (Fig. 5C and D). Importantly, we measured an age-dependent increase of ubiquitinated proteins in both single-transgenic A30P mice (A30P 4M: 159.5 ± 18.0%, A30P 18M: 387.4 ± 59.5%; P < 0.01, Fig. 5C and D) and in double-transgenic mice (double 4M: 312.9 ± 29.9%, double 18M: 516.1 ± 10.2%; P < 0.01, Fig. 5C and D). Furthermore, pairwise comparison with age-matched single-transgenic A30P mice showed an increase of ubiquitin-containing proteins in 4-months-old as well as in 18-months-old double-transgenic mice (P < 0.05) despite an equal amount of monomeric (8 kDa) ubiquitin. This result suggests that the increase in ubiquitinated proteins is not solely due to an elevated level of the free ubiquitin pool. Taken together, these findings indicate that co-expression of sph1 with A30P α-syn in double-transgenic animals led to an increased level of ubiquitin-containing proteins.
Synphilin-1 overexpression increases autophagy marker in A30P transgenic mice
Based on previous studies reporting that (i) sph1 targets misfolded proteins to aggresomes (34), (ii) α-syn and sph1 positive aggresomes are cleared by autophagy (35,36) and (iii) autophagy is induced in mice co-expressing sph1 in a human A53T α-syn transgenic mouse model (37), we analyzed expression level of a subset of autophagy markers. We first quantified the autophagic flux by measuring levels of p62, an important regulator of proteostasis and an autophagic substrate known to target ubiquitinated proteins to autophagic degradation (57). Increased levels were shown to associate with protein aggregation and impaired autophagy (58), while reduced levels reflect an increased autophagic flux (59). An increased amount of p62 (29.3 ± 13.5 and 22.0 ± 11.9%) was present in 18-months-old single-transgenic A30P mice only compared with age-matched nTg and double-transgenic mice, respectively (P < 0.05, Fig. 6A). This increase in p62 is potentially related to an impaired autophagic clearance pathway. We next analyzed beclin-1 levels, which positively correlate with the induction of phagosome nucleation (60). Importantly, we observed a strong increase in beclin-1 expression levels in 18-months-old double-transgenic animals when compared with age-matched nTg and single-transgenic A30P animals, respectively (versus nTg: 87.20 ± 29.3%: P < 0.01; and A30P: 65.7 ± 26.2% P < 0.05, Fig. 6A). We further co-stained brainstem sections, using antibodies against LAMP-2A and LC-3, implicated in the recruitment of α-syn to lysosomal degradation (27,61,62). Histologically, a punctate staining of LAMP-2A and LC-3 is apparent, corresponding to an autophagic response in single-transgenic A30P mice and in double-transgenic animals. Consistent with this finding, we observed a relative strong co-localization of A30P α-syn with autophagic markers within the soma and the perinuclear region of double-transgenic mice (Fig. 6B; arrows). Taken together, this indicates that co-expression of sph1 induces autophagy degradation and thus may lead to the clearance of misfolded A30P α-syn in aged double-transgenic animals.
Synphilin-1 overexpression is not sufficient to restore the motor deficit in aged A30P transgenic mice
Age-dependent progressive motor deficits were reported in A30P α-syn transgenic mice; initial deficits consisted of motor impairments in particular of the hind limbs, appearing between 6 and 14 months of age (42). In order to study the potential effects of sph1 on the A30P-related motor phenotype, we sequentially performed accelerating rotarod test over a 10-month-period (Fig. 7A). We observed a progressive impairment in motor coordination reflected by a decreased latency to fall from the rod. This motor deficit was first present at ∼6–7 months in single-transgenic A30P animals (nTg: 160.3 ± 10.6 s, sph1: 175.6 ± 29.7 s, A30P: 131.7 ± 18.7 s, double: 170.0 ± 22.2 sec; P < 0.05, one way ANOVA, post-hoc Bonferroni). This deficit was not related to differences in global motor skill learning (63), as mice of all groups were able to learn this task at 4 months of age (Supplementary Material, Fig. S1). Importantly, the motor phenotype was detected also in aged double-transgenic mice (Fig. 7A, inset), however, with a delay of 3 months (nTg: 142.5 ± 17.9 s, sph1: 165.8 ± 31.8 s, A30P: 84.0 ± 5.6 s, double: 104.6 ± 18.4 s; P < 0.05, one way ANOVA, post-hoc Bonferroni). In order to confirm motor deterioration in mid- and old-aged A30P and double-transgenic animals, we performed a challenging beam walk experiment (63). As expected, we observed that both single- and double-transgenic mice aged 12 months require more time to cross the beam compared with nTg control mice (nTg: 9.2 ± 1.1 s, sph1: 16.9 ± 2.3 s, A30P: 19.4 ± 5.9 s, double: 20.2 ± 4.7 s; P < 0.05, one-way ANOVA, post-hoc Bonferroni, Fig. 7B). To assess general locomotor and exploratory activity in single-transgenic A30P and double-transgenic mice, animals were subjected to an automated cage system consisting of an enriched environment, which includes a jumpable box (PhenoTyper, Noldus Information Technology) (Supplementary Material, Fig. S2A). Consistently, a strong reduction in home cage activity was observed in both single-transgenic A30P and double-transgenic mice compared with nTg mice during the dark phase (nTg: 50990 ± 8000 cm, sph1: 24116 ± 4100 cm, A30P: 34203 ± 3200 cm, double: 27292 ± 2700 cm; P < 0.05, one way ANOVA, post-hoc Bonferroni; Supplementary Material, Fig. S2B). Secondly, less exploratory and locomotor activity of transgenic mice reflected by the decreased amount of time spent on the jumpable box was noticed (using add-on of Noldus' EthoVision HeatMap Generator; Supplementary Material, Fig. S2C). To conduct a more detailed analysis regarding the underlying gait pattern, we performed an automated gait analysis using the CatWalk gait analysis system of Noldus Information Technology (Fig. 7C and D). This system gives a detailed read-out of gait parameters following video capture and manual segmentation of step patterns. It allows measurement of the number of steps and additionally the temporal dynamics of step cycles, defined by the stand (contact between paw and glass plate) and its consecutive swing phase (no contact). In aged single-transgenic A30P mice, we observed that a significant higher number of steps were necessary for crossing the walkway (nTg: 19.0 ± 1.5 steps, sph1: 19.0 ± 0.6 steps, A30P: 21.31 ± 1.4 steps, double: 20.0 ± 1.8 steps; P < 0.05, one Way ANOVA, post-hoc Bonferroni; Fig. 7C). Additionally, an increase of the stand phase in animals overexpressing sph1 and in double-transgenic animals was noticed (Fig. 7D), reflecting a reduced stability (nTg: 0.11 ± 0.004 s, sph1: 0.14 ± 0.009 s, A30P: 0.12 ± 0.006 s, double: 0.13 ± 0.003 s; P < 0.05, one Way ANOVA, post-hoc Bonferroni). These data suggest that albeit delayed, sph1 co-expression does not rescue the progressive nature of the motor phenotype in A30P animals.
The present study reveals that the long-term co-expression of sph1 with A30P α-syn in double-transgenic animals mildly mitigates the locomotor phenotype only, but leads to a profound reduction of distinct protein aggregation markers frequently observed in the brainstem of aged single-transgenic A30P mice. Interestingly, these structural and functional observations were paralleled by a strong decrease of aggregation-prone species of human α-syn, such as C-terminally truncated α-syn or the accumulation of cytosolic soluble and filamentous α-syn. Further, both transgenic proteins co-localized in inclusion bodies accompanied by an increase of ubiquitin-containing proteins as well as of autophagy markers in brainstem neurons of double-transgenic mice. Although we detected increased ubiquitin-positive protein inclusions in both single- and double-transgenic mice (Fig. 5), only those of double-transgenic mice resembled aggresomes showing both a perinuclear localization and co-localization with γ-tubulin (Figs 2D and E, 5A, 6B). The co-expression of sph1 could potentially influence autophagosomal processing and development of aggresomes on several levels. First, sph1 interferes with the degradation of mutant A30P α-syn per se: protein–protein interaction between sph1 and α-syn was reported to slow down α-syn degradation by the proteasomal pathway in vitro (31). Aggresomes may serve as a structural prerequisite for the disposal of protein aggregates resistant to proteasomal degradation via the autophagic route (35). Indeed, several studies demonstrated that autophagy- and proteasome-based degradation pathways underlie a potential cross-talk (28,64–66). For example, the work from Ebrahimi-Fakari et al. (28) suggests that raised intracellular α-syn levels results in higher activation of UPS degradation pathways, when autophagy is inhibited. This may apply to the single-transgenic A30P mice, since the A30P mutation was reported to impair autophagy-mediated clearance (67). Further, we consistently observed a selective increase in p62 in single-transgenic A30P mice (Fig. 6A), which is known to accumulate in inclusion bodies associated with impaired autophagy (57). Additionally, α-syn degradation by the UPS resulted in an accumulation of truncated species in vitro (68,69) promoting formation of α-syn fibrils (68). We observed an increase in N- and C-terminally truncated α-syn species and fibrils in single-transgenic A30P mice only (Fig. 1D). Importantly, sph1 interacts with the N- and C-terminal region of human wild-type and A30P α-syn (70) potentially inhibiting its degradation by the proteasome (31). Thus, an increase in truncated α-syn species, eventually caused by its predominant clearance via the UPS pathway, may alter the physiological interaction of both proteins promoting different degradation pathways in single- and double-transgenic mice, and possibly also contributing to PD. Accordingly, the profound reduction of truncated α-syn, as well as thioflavin S- and PK-resistant α-syn fibrils suggests that A30P α-syn co-localizing within aggresomal structures might be predominantly targeted by autophagic clearance pathways in the present double-transgenic mice.
Secondly, beclin-1 upregulation detected in double-transgenic mice may promote the formation of autophagosomal complexes. In this regard, overexpression of beclin-1 restored impaired autophagy not only in beclin-1 knockout mice (71), but also in several other animal models displaying decreased autophagic clearance of misfolded proteins (72,73). Similarly, increased beclin-1 levels were also reported in mice overexpressing A53T α-syn and sph1 (37), however, without specifying the expression level of distinct α-syn species, a novel finding of the present study. Although the underlying mechanisms leading to beclin-1 upregulation are not yet fully understood, its increase further indicates a predominant role of autophagy in the reduction of aggregation marker in double-transgenic mice.
Thirdly, sph1 binds to lipid rafts in endomembranes forming aggresomes upon approaching the diauxic shift in yeast (74). This process may recruit α-syn into aggresomal structures, possibly transferred via retrograde transport along microtubules to the perinuclear region (49,75). While single-transgenic A30P mice display accumulation of aggregation marker both in cell bodies and neuropil (Figs 2C, 4B and C), α-syn and ubiquitin-positive deposits were predominantly localized within the perinuclear region in double-transgenic mice (Figs 2D and E, 5A, 6B). Indeed, both A30P mutant α-syn as well as α-syn aggregates were shown to interact with Rab proteins involved in endolysosomal trafficking and axonal retrograde transport (76–78). It is therefore possible, that co-expression of sph1 restores trafficking function in double-transgenic mice.
On the other hand, mutant A53T α-syn was also detected in aggresome-like formations in A53T and sph1 double-transgenic mice (37). However, the A53T α-syn aggregation pathology is more frequently observed within neurites and/or synapses (53,79,80), whereas A30P aggregates are formed within the cell body, possibly related to its decreased binding capacity towards synaptic membranes (81,82). These findings suggest that co-expression of sph1 could induce sequestration of the mutant protein also prior to its transport to the synapse. This may explain the presence of either transgenic protein in perinuclear aggresome-like structures of both double-transgenic mouse lines.
Note, although the formation of aggresomes is usually considered to be cytoprotective (32,37), sph1 can also form detrimental aggresomes (74,83) and their occurrence is often associated with deregulation of normal cellular homeostasis (75).
Additionally, the functional phenotype, although delayed, was essentially similar for aged double- and single-transgenic mice despite its profound difference in fibril and aggresome formation. Consistently, the progressive nature of motor deterioration, as measured by several behavioral tests, astrocytic activation, ubiquitination and insolubility were still present in aged mice and may contribute to the underlying neuropathology. The detected differences in gait pattern may point to additional endogenous compensatory mechanisms of neural circuits governing motor function, which, however, remain to be investigated in more detail in future studies.
Taken together, our data demonstrate that sph1 promoted aggresome formations potentially susceptible to autophagic clearance, strongly reducing aggregation-prone truncated and fibrillized α-syn species. However, this was insufficient to inhibit α-syn conversion into insolubility and to fully restore motor dysfunction. Hence our findings are also in favor of accumulating evidence, that fibrillized α-syn is not the sole neurotoxic entity in PD progression.
MATERIAL AND METHODS
Generation of double-transgenic mice overexpressing human synphilin-1 on a human A30P alpha-synuclein overexpressing background
The generation of transgenic mice overexpressing human A30P α-syn under the mouse thymus cell antigen 1 (Thy-1) promoter, as well as the generation of transgenic mice overexpressing human sph1 under the mouse prion protein (PrP) promoter, have been described previously (39,40,41). Double-transgenic mice (double) were generated by crossbreeding α-syn transgenic mice (A30P) to synphilin-1 transgenic mice (sph1). All experiments were approved by German animal welfare authorities.
Order of the behavioral tests was determined in regard of stress induced to the animal (84). Between each tests, a pause of at least 48 h was applied. The experimenter was blind to the genotype of each mouse. Animals were allowed a minimum of 15 min to acclimate to the experimental room prior testing procedures. Behavioral group was composed of 12 mice, and general animal health was screened using SHIRPA observations prior conductance of behavioral analyses.
To assess motor function and coordination in voluntarily walking mice, automated gait analysis was performed (CatWalk XT 8.0; Noldus, Wageningen, The Netherlands). Eighteen-month-old animals were trained to cross a horizontal glass runway illuminated using LED light and reflecting footprint captured by a high-speed video camera positioned underneath the walkway. Five straight tracks without any interruption were analyzed per animals.
Challenging beam traversal
Motor performance was measured with a challenging beam traversal test as previously described (85). Briefly, 12-months-old animals were trained to traverse a Plexiglas beam consisting of four sections (25 cm each, 1 m total length) of different width (3.5–0.5 by 1 cm increments) leading to the animals' home-cage. After 2 days of training, a mesh grid of 1 cm2 was placed 1 cm over the beam surface and time to traverse the beam was measured.
Accelerating rotating rod
Motor coordination and motor skill learning were evaluated using an accelerating rotarod (TSE rotarod Systems; TSE, Bad Homburg, Germany) and time spent on the rod was recorded. Mice were pre-trained on rotarod 2 days prior to the test. The first training day consisted of three trials at constant speed (4, 10 and 20 rpm for 5 min) and the second training day was a combination of two constant speed runs (10 and 20 rpm for 5 min) and a progressive accelerating speed ranging from 4 to 40 rpm in the time course of 7 min. Mice were then tested on a 4–40 rpm progressive acceleration within 7 min for five consecutive days including three trials each day. An inter-trial pause of at least 1 h was applied to avoid fatigue and stress. Motor coordination was evaluated by comparing the mean latency to fall over five consecutive days between groups. Motor skill learning was estimated by the mean of performance over the first testing week in 4-months-old mice.
Automated home cage
Analysis of home cage activity was investigated using an automated tracking system of Noldus Information Technology (PhenoTyper system, Noldus: L = 30 × W = 30 × H = 35 cm). Briefly, home cage locomotor activity was analyzed in eight singly housed mice of each group over a 3-week-period. Experiment was performed under a 12 h light and 12 h dark cycle. A jumpable shelter (L = 10 × W = 10 × H = 5 cm) was fixed in one of the corners. Home cage locomotor activity was investigated by continuous monitoring of mice using EthoVision XT (Noldus Information Technology, Noldus) and data gained from the third testing week were analyzed using MatLab (MathWorks, Natice34qwk, MA, USA).
Sequential protein extraction
Mice were sacrificed by asphyxiation using increasing concentrations of CO2. Next, brain tissues were freshly prepared on an ice cold plate, immediately snap frozen in liquid nitrogen and stored at −80°C. Mouse brain was cut coronally approximately at bregma −3 mm to separate forebrain and hindbrain region. Sequential protein extraction was performed using buffers with different salinity and detergent concentration followed by intermittent ultracentrifugation steps at 120 000g as described previously (50) with minor modifications. Briefly, homogenization of tissue was performed at 30 000 rpm for 30 s using a tissue homogenizer (Ultra-Turrax; IKA Werke, Staufen, Germany) in five volumes of Tris-buffered saline (TBS, pH 7.5, 10 mm Tris, 0.15 M NaCl) supplemented with a cocktail of protease inhibitors (Complete, Roche Applied Science) and centrifuged (30 min, 120 000g, 4°C). Resulting supernatant represented the TBS soluble fraction. Pellet was then extracted sequentially using TBS containing 1% of Triton X-100 (Sigma-Aldrich) and protease inhibitors, TBS containing 0.1 M sucrose and protease inhibitors, RIPA buffer (50 mm Tris–HCl, pH 7.4, 175 mm NaCl, 5 mm EDTA, 1% NP-40, and 0.5% sodium deoxycholate, 0.1% SDS). Finally, the detergent-insoluble pellet was solubilized in 2 m urea/5% SDS (86).
Western blot analysis and antibodies
Protein concentration was measured using the Bradford method (Protein Assay Dye Reagent Concentrate; Biorad; Munich, Germany). Protein lysates were mixed with loading buffer (80 mm Tris, pH 6.8, 0.1 mm DTT, 2% SDS, 10% glycerol, bromophenol blue), denatured by heating at 95°C for 5 min, and analyzed in polyacrylamide gel electrophoresis (PAGE) buffer (0.2 m glycine, 25 mm Tris, 1% SDS) by SDS–PAGE (Blue Vertical 100/C; Serva, Heidelberg, Germany) via a 8–16% gradient gel (Serva). Gels were stabilized in transfer buffer containing 15% methanol and transferred onto a nitrocellulose membrane (Whatman, Dassel, Germany) or a methanol-activated PVDF (Immobilon P; Millipore Corporation, Billerica, MA, USA) for 2 h at 80 V. Blots were subsequently blocked for 1 h with 5% dry milk in TBST buffer (10 mm Tris, pH 7.5, 0.15 m NaCl, 0.1% Tween 20) and then incubated for 2 h with a primary antibody in TBST containing 5% dry milk. Mouse anti-α-syn Mc42 antibody raised against amino acids 91–99 in the NAC domain (610787; 1:3.000, Transduction Laboratories, Lexington, KY, USA) was used to detect both, mouse and human α-syn and rat anti-α-syn 15G7 antibody raised against amino acids 116–131 of the C-terminal [1:50] (41) was used to specifically visualize human α-syn. Rabbit anti-ubiquitin (Z0458; 1:1.000, Dako Cytomation, Denmark) and rabbit anti-ubiquitin (MAB1510, 1:1.000, Millipore) were used to detect ubiquitinated proteins. Beclin-1 (1:100, Novus Biologicals, Littleton, CO, USA) and p62 (1:200, BD Bioscience, San Diego, CA, USA) were used to detect autophagy marker. Mouse anti-beta-actin antibody (A4700; 1:10.000, Sigma) was used as internal loading control. After washing with TBST, the membrane was incubated for 75 min with a secondary antibody coupled to horseradish peroxidase (GE Healthcare; Piscataway, NJ, USA) diluted in TBST containing 5% dry milk. Protein bands were detected using chemiluminescence method (ECL; GE Healthcare) and exposure to film (Hyperfilm ECL, GE Healthcare). For densitometric analysis, films were scanned and densitometry was performed using imageJ [NIH] (87). Statistical analysis was performed using the ratios of the densitometric value of each band and its corresponding beta-actin loading control within each genotype group.
Samples were diluted 1:1 into non- reducing Laemmli sample buffer (150 mm Tris–HCl pH 6.8, 33% glycerol, 1.2% SDS and bromophenol blue), loaded on 1.5% agarose gel (Biorad), 375 mm Tris–HCl pH 8.8, 0.1% SDS), and incubated for 5 min at 95°C. 0.15 mg of total protein was loaded per AGERA lane. After loading, gels were run in Laemmli running buffer (192 mm glycine, 25 mm Tris-base, 0.1% SDS) at 120 V. Gels were blot on PDVF membranes (Immobilon-P; Millipore) at 80 V for 2 h (Bluepower 500; Serva) on transfer buffer (192 mm glycine, 25 mm Tris-base, 0.1% SDS, 15% methanol) and blocked for 2 h at room temperature in 5% dry milk in TBST buffer.
Immunoelectron microscopy of purified filamentous microaggregates
Isolation of filamentous microaggregates from mouse brain tissue was performed as described previously (88). Briefly, brains were homogenized in buffer 1 (10 mm Tris, 1 mm EGTA, 0.8 m NaCl, 10% sucrose, 0.1% Triton X-100 and protease inhibitors) in a glass homogenizer. After centrifugation (Sorvall SS34 rotor; Thermo Scientific, Rockford, IL, USA) for 20 min at 2.988g, the supernatant S1 was saved and pellet P1 was homogenized in buffer 1 and centrifuged again. The supernatants S1 and S2 were combined, adjusted to 1% N-lauroylsarcosine and 1% β-mercaptoethanol and incubated for 2 h at room temperature. After centrifugation (TLA-100.3 rotor; Beckman Coulter, Fullerton, CA, USA) for 45 min at 134.877g, filament-containing pellets were homogenized in buffer 2 (10 mm Tris, 1 mm EGTA, 0.8 M NaCl, 10% sucrose and protease inhibitors), layered over a discontinuous sucrose gradient consisting of 50% sucrose and 20% sucrose in buffer 3 (10 mm Tris, 1 mm EGTA and 0.8 m NaCl) and centrifuged (SW40 rotor; Beckman Coulter) for 2 h at 217.290g. Then, the different fractions were collected and stored at −70°C until use.
Immunoelectron microscopy for α-syn was performed and quantified as described before (44). Briefly, the fractions obtained from the sucrose gradient were adsorbed to electron microscopy carbon-coated grids. Then, samples were incubated with rabbit anti-α-syn (C-20; Santa Cruz Biotechnology, Santa Cruz, CA, USA). Grids were then incubated with protein A-gold (10 nm; Sigma). Finally, the samples were counterstained with 2% uranyl acetate. Transmission electron microscopy was performed in 1200EX electron microscope (Jeol, Peabody, MA, USA) model operated at 100 kV. Overview of the samples revealed that the 10–20% interface fraction was the most enriched in α-syn positive filament and, accordingly, counting for comparison among the different transgenic mouse lines was performed on 10–20% interface fractions. For α-syn filament counting, grids for each condition were analyzed by counting 250 µm2 per grid at ×15 000 magnification.
Histological and immunohistological analysis
Mice were deeply anesthetized with carbon dioxide and transcardially perfused with phosphate-buffered saline (PBS, pH 7.5, 137 mm NaCl, 2.7 mm KCl, 4.3 mm Na2HPO4 · 7H2O, 1.4 mm KH2PO4, room temperature) and 4% cold paraformaldehyde (PFA, pH 7.5, 4°C) diluted in PBS. Brains were removed, postfixed overnight and embedded in paraffin. Fixed brains were sagittally cut into 7 µm sections. Deparaffinization of slides was performed in xylene and rehydration was made using solutions containing a decrease in ethanol concentration. For detection of proteinase K (PK)-resistant α-syn, sections were additionally incubated in digestion buffer (10 mm Tris/HCl pH 7.8, 100 mm NaCl, 0.1% Nonidet p-40) containing 50 µg/ml PK for 30 min at 37°C. Next, slides were microwaved for 15 min in 10 mm sodium citrate (pH 6.0) and washed with PBS. Endogenous peroxidase was blocked using 0.3% hydrogen peroxide (H2O2). After blocking (5% normal serum, 0.3% Triton X-100 in PBS), sections were washed with PBS, and primary antibody (diluted in PBS, 1% normal serum) was incubated overnight. After washing in PBS, secondary antibody coupled with biotin (Vector Laboratories; Burlingame, CA, USA) was diluted 1:200 in PBS and 1% normal serum, and incubated for 1 h. After washing with PBS, an avidin–biotin enhancer complex coupled with peroxidase (ABC Elite; Vector Laboratories) was added and incubated at room temperature for 2 h. After washing with PBS, 3,3′-diaminobenzidine (DAB; Sigma) was added on slices, and reaction was stopped in distilled water. Slides were counterstained using hematoxylin (Merck; Darmstadt, Germany), dehydrated and mounted (CV mount medium; Leica, Bensheim, Germany). For light microscopical analysis, sections were viewed using an Axioplan2 imaging microscope (Carl Zeiss; Oberkochen, Germany) equipped with an AxioCam HR color digital camera (Zeiss) and the AxioVision4.3 software package (Zeiss). Cell counting was performed by a blinded unbiased method, using three mice per group, four slices per mice and eight non-overlapping counting frames per slices. Counting frames were taken at ×20 magnification (270 × 200 µm) and disposed in the pontine reticular formation of the brainstem. Staining events and nuclei were evaluated using ImageJ. The triangle thresholding method and the analysis particles plugin were used and additionally confirmed by using a manual counting. Values obtained were normalized by total number of nuclei per field.
Immunofluorescence and thioflavin S analysis
7-µm-sagittal sections were deparaffinized in xylene and blocked for 1 h (10% normal serum, 0.3% Triton X-100 in TBS). Sections were incubated with primary antibody (diluted in TBS, 1% normal serum, 4°C) overnight. Rat anti-α-syn 15G7 antibody (1:25) was used to visualize human α-syn and rabbit anti-myc (1:500) was used to detect the N-terminal myc-tagged transgenic sph1. Mouse anti-γ-tubulin antibody (T6557, 1:100, Sigma) was used to detect centrosome and to validate aggresome formation. Co-staining of human anti-α-syn using the 15G7 antibody (1:25) and LAMP-2A (1:200, Abcam, ab18528) and LC3 (2G6, 1:500, nanoTools), respectively, were used to detect colocalization of α-syn with lysosomal structures. After washing with TBS, secondary antibodies coupled to a fluorescent dye (DyLight; Thermo Scientific) were incubated at room temperature for 1 h. After washing, nuclei were counterstained using 4′,6-diamidino-2-phenylindole (DAPI; Invitrogen, Carlsbad, CA, USA) diluted 0.5 µg/ml in TBS for 10 min, washed and coverslipped using aqueous mounting media (Mowiol, Calbiochem, San Diego, CA, USA) and anti-fade for fluorescence microscopy (DAPCO, 25 mg/ml, Sigma).
Thioflavin S staining was performed after deparaffinization in xylene and hydratation of sections, followed by 10 min exposure to 0.1% thioflavin S (Sigma). Subsequently, slides were washed in 70% ethanol and then rinsed in distilled water and coverslipped.
This work was supported by a European Commission Marie Curie Initial Training Network Grant Agreement: 215618 to O.R.; Federal Ministry of Education and Research (01GN0979), the Albert-Raps Foundation, grants of the University Hospital Erlangen (ELAN No. 08.11.05.1; IZKF No. TP9), Bavaria California Technology Center (BaCaTeC), the Bavarian State Ministry of Sciences, Research, and the Arts, ForNeuroCell, the Spanish Ministry of Science and by Fundación Ramón Areces. S.N. is a fellow of the German Parkinson Society.
We thank D. Berg for expert advice, M. Diepenbroek and J. Weber for critical reading. We also thank M. Mellace for his intensive help with automated cages. We further thank H. Esmer, M. Yesilyurt, S. Mayer, A. Hoffmann, A. Mueller and W. Yiwen for excellent technical help and J. Winkler for valuable discussion and comments on the manuscript.
Conflict of Interest statement. None declared.