Synapsin III gene silencing redeems alpha-synuclein transgenic mice from Parkinson's disease-like phenotype

Abstract

Fibrillary aggregated α-synuclein (α-syn) deposition in Lewy bodies (LB) characterizes Parkinson's disease (PD) and is believed to trigger dopaminergic synaptic failure and a retrograde terminal-to-cell body neuronal degeneration. We described that the neuronal phosphoprotein synapsin III (Syn III) cooperates with α-syn to regulate dopamine (DA) release and can be found in the insoluble α-syn fibrils composing LB. Moreover, we showed that α-syn aggregates deposition, and the associated onset of synaptic deficits and neuronal degeneration occurring following adeno-associated viral vectors-mediated overexpression of human α-syn in the nigrostriatal system are hindered in Syn III knock out mice. This supports that Syn III facilitates α-syn aggregation. Here, in an interventional experimental design, we found that by inducing the gene silencing of Syn III in human α-syn transgenic mice at PD-like stage with advanced α-syn aggregation and overt striatal synaptic failure, we could lower α-syn aggregates and striatal fibers loss. In parallel, we observed recovery from synaptic vesicles clumping, DA release failure, and motor functions impairment. This supports that Syn III consolidates α-syn aggregates, while its downregulation enables their reduction and redeems the PD-like phenotype. Strategies targeting Syn III could thus constitute a therapeutic option for PD.

Keywords: Parkinson's disease; alpha-synuclein; dopamine release; dopaminergic neurons; motor functions; synapsin III; synaptic vesicles.

Graphical abstract

Figure 1

Evaluation of the efficiency and specificity of Syn III silencing in the nigrostriatal system of AAV-shNSC- and AAV-shSynIII-injected mice (A) Representative images showing eGFP-immunolabeling in the substantia nigra and striatum of SYN120 tg mice injected with AAV-shNSC or AAV-shSynIII. Scale bar: 100 μm. (B) Representative images show eGFP signal in the TH-immunopositive neurons of the substantia nigra and fibers of the striatum of AAV-shNSC- and AAV-shSynIII-injected SYN120 tg mice. Scale bar: 20 μm. (C) Syn III-immunopositive signal in the TH-positive neurons of the substantia nigra of C57BL/6J, C57BL/6JOlaHsd, and SYN120 tg mice injected with either AAV-shSNC or AAV-shSynIII. Graphs show results from the analysis of the Syn III-positive area in the substantia nigra of AAV-shSynIII-injected mice when compared with the AAV-shNSC-injected littermates of the three mouse lines. Data are expressed as positive area in μm². Boxplots represent the distribution of the 75%, 50%, and 25% of the values. Whiskers indicate the upper and lower extreme of the dataset. Please note the significant decrease of Syn III in AAV-shSynIII-injected mice (∗∗∗, –2,221 μm², P < 0.001 in C57BL/5J mice; ∗∗∗, –2,081 μm², P < 0.001 in C57BL/6JOlaHsd mice; ∗∗∗, –3,057 μm², P < 0.001 in SYN120 tg mice). Unpaired Welch's t test. n = 4/5 for each group. Scale bar: 20 μm.

Figure 2

Efficiency of Syn III gene silencing in the AAV-shSynIII-injected C57BL/6J, C57BL/6JOlaHsd, and SYN120 tg mice (A) Representative images show Syn III and DAT-immunolabeling in the striatum. Graphs in middle panels show total Syn III-positive area (in μm²) in the striatum of AAV-shSynIII-injected-mice when compared with the AAV-shNSC-injected mice. A significant reduction of Syn III-positive area in the AAV-shSynIII-injected C57BL/5J (∗∗∗, –695 μm², P < 0.001), C57BL/6JOlaHsd (∗∗∗, –942 μm², P < 0.001), and SYN120 tg mice (∗∗∗, –1942 μm², P < 0.001) was detected. Graphs in right panels show the average size of Syn III-positive areas. Only SYN120 tg mice exhibited a significant decrease in this parameter (∗∗∗, –0.33 μm², P < 0.001). Boxplots represent the distribution of the 75%, 50%, and 25% of the values. Whiskers indicate the upper and lower extreme of the dataset. Unpaired Welch's t test. n = 4/5 for each group. Scale bar: 20 μm. (B) Graphs show results of qRT-PCR results assessing the relative expression of Syn III, Syn II, and Syn I mRNA (expressed as fold changes in mRNA expression versus the mean mRNA levels of AAV-shNSC-injected C57BL/6J mice) in the substantia nigra of AAV-shSynIII- or AAV-shNSC-injected C57BL/6J, C57BL/6JOlaHsd, and SYN120 tg mice. A significant decrease of Syn III mRNA levels in the AAV-shSynIII-injected C57BL/6J (∗, –0.56, P < 0.05), C57BL/6JOlaHsd (∗, –1.93, P < 0.05), and SYN120 tg (∗, –2.41, P < 0.05) animals with respect to their AAV-shNSC-injected littermates was detected. Syn III mRNA expression was increased in AAV-shNSC-injected C57BL/6JOlaHsd and SYN120 tg mice when compared to the C57BL/6J animals (∗, +2.38, P < 0.05; ∗, +2.66, P < 0.05). No changes in Syn II mRNA expression were detected, while only SYN120 tg mice exhibited a marked increase in Syn I mRNA levels after the AAV-shSynIII injection when compared with the AAV-shNSC-injected littermates (∗∗, +5.21, P < 0.01), with AAV-shNSC- and AAV-shSynIII-inoculated C57BL/6J mice (∗∗, +5.76, P < 0.01; ∗∗∗, +5.27, P < 0.001) and with AAV-shNSC-injected C57BL/6JOlaHsd (∗∗, +5.10, P < 0.01). Data are expressed as mean ± SEM. Two-way ANOVA + Bonferroni's multiple comparisons test. n = 4/5 for each group.

Figure 3

Western blot analysis evaluating the efficiency and specificity of Syn III silencing in the substantia nigra and striatum of AAV-shNSC- and AAV-shSynIII-injected mice (A) Syn III protein levels were reduced in the AAV-shSynIII-injected substantia nigra of C57BL/6J (∗, –0.34, P < 0.05), C57BL/6JOlaHsd (∗, –0.46, P < 0.05), and SYN120 tg mice (∗, –0.52, P < 0.05) when compared to AAV-shNSC-injected littermates. Please note the statistically significant increase of Syn III in C57BL/6JOlaHsd (∗∗, +0.62, P < 0.01) and SYN120 tg mice (∗∗, +0.63, P < 0.01) when compared to AAV-shNSC-injected C57BL/6J mice. No difference in Syn II or Syn I levels was observed. Data are expressed as mean ± SEM. Two-way ANOVA + Bonferroni's multiple comparisons test. n = 6 mice for each group. (B) The analysis of Syn III levels in the striatum showed a significant decrease in the AAV-shSynIII-injected C57BL/6J (∗, –0.29, P < 0.05), C57BL/6JOlaHsd (∗, –1.01, P < 0.05), and SYN120 tg mice (∗, –1.07, P < 0.05) when compared to AAV-shNSC-injected littermates. Please note the statistically significant increase of Syn III in C57BL/6JOlaHsd (∗, +0.82, P < 0.05) and SYN120 tg mice (∗∗, +0.92, P < 0.01) when compared to AAV-shNSC-injected C57BL/6J mice. No differences were observed in Syn II or Syn I levels following Syn III gene silencing in the three mouse lines analyzed. Data are expressed as mean ± SEM. Two-way ANOVA + Bonferroni's multiple comparisons test. n = 6 mice for each group.

Figure 4

Gene silencing of Syn III prevents α-syn aggregation (A) Representative images of thioflavin-S staining on α-syn-immunolabeled sections from the substantia nigra of AAV-shNSC or AAV-shSynIII-injected SYN120 tg mice. The graph shows the significant decrease of thioflavin-S-positive area in the animals exposed to AAV-shSynIII when compared to AAV-shNSC littermates (∗∗∗, –161 μm², P < 0.001). Boxplots represent the distribution of the 75%, 50%, and 25% of the values. Whiskers indicate the upper and lower extreme of the dataset. Unpaired Welch's t test. n = 5 mice for each group. Scale bar: 20 μm. (B) Representative immunolabeling of α-syn coupled with thioflavin-S staining in the striatum of AAV-shNSC- or AAV-shSynIII-injected SYN120 tg mice. The graph is showing the significant decrease of thioflavin-S-positive area of AAV-shSynIII when compared to AAV-shNSC littermates (∗∗∗, –136 μm², P < 0.001). Boxplots represent the distribution of the 75%, 50%, and 25% of the values. Whiskers indicate the upper and lower extreme of the dataset. Unpaired Welch's t test. n = 5 mice for each group. Scale bar: 20 μm. (C) Western blot analysis of the striatal UREA/SDS soluble fractions of AAV-shNSC- and AAV-shSynIII-injected SYN120 tg mouse brains revealed a statistically significant decrease of both monomeric and high molecular weight α-syn in the mice subjected to Syn III gene silencing (∗∗∗, –0.98, P < 0.001 and ∗, −3.65, P < 0.05). Data are expressed as mean ± SEM. Unpaired Welch's t test. n = 4 mice for each group. (D) Western blot analysis of the striatal RIPA-soluble fractions of AAV-shNSC- and AAV-shSynIII-injected SYN120 tg mouse brains revealed no differences in the monomeric levels of α-syn. Data are expressed as mean ± SEM. Unpaired Welch's t test. n = 4 mice for each group. (E) Analysis of the spectral emission wavelength of the luminescent-conjugated oligothiophene HS-68 in striatal sections exposed to sequential α-syn-immunolabeling revealed with an Alexa 647 secondary antibody and HS-68 labeling. The SYN120 tg mice injected with AAV-shNSC exhibited a higher fluorescence intensity when compared to the AAV-shSynIII-injected littermates. n = 3 animals for each group. (F) The ratio between the peaks at λ = 485 nm and λ = 570 nm showed that the SYN120 tg mice injected with AAV-shNSC showed a blue shifted spectrum when compared the SYN120 tg mice exposed to Syn III gene silencing. Data are expressed as mean ± SEM. ∗P < 0.05, Unpaired Welch's t test. n = 3 animals for each group.

Figure 5

Syn III gene silencing reduces dopaminergic fiber loss in the striatum of SYN120 tg mice (A) Representative images of TH-positive fibers in the striatum of AAV-shNSC- or AAV-shSynIII-injected C57BL/6J, C57BL/6JOlaHsd, and SYN120 tg mice. Graph shows densitometric analysis of TH-positive area (in μm²) in the striatum of AAV-shNSC- or AAV-shSynIII-injected C57BL/6J, C57BL/6JOlaHsd, and SYN120 tg mice. AAV-shNSC-injected SYN120 tg mice showed a significant reduction of TH-positive area when compared to C57BL/6J (∗∗∗, –355 μm², P < 0.001 for AAV-shNSC-injected mice; ∗∗∗, –351 μm², P < 0.001 for AAV-shSynIII-injected mice), and C57BL/6JOlaHsd (∗∗∗, –345 μm², P < 0.001 for AAV-shNSC-injected mice; ∗∗∗, –307 μm², P < 0.001 for AAV-shSynIII-injected mice). TH-positive fiber loss in SYN120 tg mice was recovered by Syn III gene silencing (∗∗∗, +242 μm², P < 0.001). Data are expressed as mean ± SEM. Two-way ANOVA + Bonferroni's multiple comparisons test. n = 6 for each group. Scale bar: 50 μm. (B) Western blot analysis showed that TH levels (expressed as optical density, O.D.) were reduced in the striatum of AAV-shNSC-injected SYN120 tg mice when compared to C57BL/6J (∗, –0.32, P < 0.05) or C57BL/6JOlaHsd animals (∗, –0.3, P < 0.05 for AAV-shNSC-injected mice; ∗, –0.47, P < 0.05 for AAV-shSynIII-injected mice). TH levels were significantly higher in the AAV-shSynIII-injected SYN120 tg mice when compared to AAV-shNSC-injected SYN120 tg littermates (∗∗, +0.24, P < 0.01). Data are expressed as mean ± SEM. Two-way ANOVA + Bonferroni's multiple comparisons test. n = 6 for each group. Asterisk indicates a non-specific band that did not contribute to the quantification. (C) Graph shows qRT-PCR results for the relative TH expression (as fold changes of mRNA expression versus the mean mRNA levels of AAV-shNSC-injected C57BL/6J mice) in the substantia nigra of C57BL/6J, C57BL/6JOlaHsd, and SYN120 tg mice. In SYN120 tg mice TH mRNA expression was significantly increased by Syn III silencing (∗, +2.03, P < 0.05). In the AAV-shNSC-injected SYN120 tg animals, TH expression was significantly lower than in C57BL/6J (∗, –1.18, P < 0.05 for AAV-shNSC-injected mice; ∗, –1.21, P < 0.05 for AAV-shSynIII-injected mice), or C57BL/6JOlaHsd controls (∗, –0.94, P < 0.05). Data are expressed as mean ± SEM. Two-way ANOVA + Bonferroni's multiple comparisons test. n = 4/5 for each group.

Figure 6

Syn III gene silencing reduces SV clustering in SYN120 tg mice (A) Representative electron micrographs of pre-synaptic terminals for each experimental group. Scale bar: 100 nm. Image on the right is annotating the SVs (in light blue), the active zone (in red), the post-synaptic density (in green) within the presynaptic terminal (in yellow). (B) Quantification of vesicle density expressed as number of vesicles per synapse area (nm²) in each experimental group. C57BL/6J AAV-shNSC: 60 synapses; C57BL/6JAAV-shSynIII: 60 synapses; C57BL/6JOlaHsd AAV-shNSC: 59 synapses; C57BL/6JOlaHsd AAV-shSynIII: 60 synapses; SYN120 tg AAV-shNSC: 60 synapses; SYN120 tg AAV-shSynIII: 63 synapses. Data are presented as mean ± SEM and analyzed by Kruskall-Wallis non-parametric test with Dunn's multiple comparison test (∗∗P < 0.01, ∗∗∗∗P < 0.0001). (C–H) Analysis of vesicle clustering by multiple-peak Gaussian fit of the frequency distribution of vesicle clustering in each experimental group in striatal synapses of C57BL/6J mice (C and D), C57BL/6JOlaHsd mice (E and F), and SYN120 tg mice (G and H). In the graphs, the red peaks represent population 1, which include the SV fitting under the curve with the median nearest neighbor distance (NND) = 34.05 nm, while the green peaks represent population 2, which includes the SV fitting under the curve with a median NND of 43.34 nm. The blue line indicates the cumulative fit of the NND of the two populations of SVs. Data are expressed in nanometers as frequency distribution of NND and fitted by multiple-peak Gaussian fit, comparing the AAV-shNSC and AAV-shSynIII conditions in each genotype. (I) The table shows the ratio between the area under the peak of population 1 and of population 2 and the respective 95% confidence interval (CI) of the AAV-shNSC and AAV-shSynIII-injected C57BL/6J, C57BL/6JOlaHsd, and SYN120 tg mice.

Figure 7

Syn III gene silencing abolished motility deficits in SYN120 tg mice (A) Time spent to travel the beam by AAV-shNSC- or AAV-shSynIII-injected C57BL/6J, C57BL/6JOlaHsd, and SYN120 tg mice. The AAV-shNSC SYN120 tg mice spent more time than C57BL/6JOlaHsd (∗, +4.06 s, P < 0.05) or C57BL/6J controls (∗∗, +7.57 s, P < 0.01). Syn III deletion reduced the beam traveling time in SYN120 tg mice (∗∗∗, –7.03 s, P < 0.001). One-way ANOVA + Newman-Keuls multiple comparisons test. n = 10 mice for each group. (B) Mean distance traveled by mice in the open field in basal condition (left graph) or following cocaine administration (right graph). AAV-shNSC-injected SYN120 tg mice exhibited reduced basal motility versus C57BL/6J (∗, –2.64 m, P < 0.05) and C57BL/6JOlaHsd littermates (∗, –2.78 m, P < 0.05). AAV-shSynIII injected C57BL/6J and SYN120 tg mice showed increased motility versus AAV-shNSC-injected littermates (∗∗, +3.21 m, P < 0.01 and ∗∗∗, +4.33 m, P < 0.001, respectively). AAV-shSynIII-injected C57BL/6JOlaHsd mice had reduced activity versus AAV-shNSC littermates (∗, –2.31 m, P < 0.05) and AAV-shSynIII-injected SYN120 tg mice (∗∗∗, –3.86 m, P < 0.001). AAV-shSynIII injection reduced cocaine response in C57BL/6J mice (∗, –4.38 m, P < 0.05). SYN120 tg mice with AAV-shNSC injection showed decreased cocaine-induced locomotion versus C57BL/6J littermates (∗, –5.08 m, P < 0.05), but this was recovered by Syn III deletion (∗∗∗, +11.41 m, P < 0.001). C57BL/6JOlaHsd motility was not changed by cocaine administration. One-way ANOVA + Newman-Keuls multiple comparisons test. n = 8/10 mice for each group.

Figure 8

Syn III gene silencing restored striatal DA release in SYN120 tg mice (A) Striatal DA release (pg/μL) as assessed by vertical microdialysis coupled with HPLC assays in basal condition (left graph) or following K⁺ (central graph) or cocaine stimulation (graph on the right). AAV-shNSC-injected SYN120 tg mice exhibited a significant reduction in basal DA release when compared to AAV-shNSC-injected C57BL/6J (∗∗, –18.42 pg/μL, P < 0.01) and C57BL/6JOlaHsd controls (∗, –15.54 pg/μL, P < 0.05). AAV-shSynIII injection increased basal DA release in C57BL/6J (∗∗, +23.40 pg/μL, P < 0.01) and SYN120 tg mice (∗, +21.18 pg/μL, P < 0.05) when compared to their respective AAV-shNSC-injected littermates. Upon Syn III deletion, C57BL/6JOlaHsd mice exhibited a mild non-significant decrease in basal DA release. AAV-shSynIII-injected C57BL/6J showed increased K⁺-stimulated DA release versus AAV-shNSC-injected littermates (∗, +30.50 pg/μL, P < 0.05). SYN120 tg exhibited a significant decrease of DA release versus AAV-shNSC-injected C57BL/6J mice (∗, –48.87 pg/μL, P < 0.05), that was rescued by Syn III deletion (∗, +84.88 pg/μL, P < 0.05 versus AAV-shNSC SYN120 tg littermates). C57BL/6JOlaHsd had no change in K⁺-stimulated DA release following Syn III deletion. AAV-shSynIII-injected C57BL/6J mice showed reduction of cocaine-induced DA release versus AAV-shNSC littermates (∗, –44.17 pg/μL, P < 0.05). SYN120 tg mice exhibited a decrease of cocaine-stimulated DA release when compared to AAV-shNSC-injected C57BL/6J animals (∗, –44.17 pg/μL, P < 0.05), but this was rescued by Syn III gene silencing (∗, +21.11 pg/μL, P < 0.05 versus AAV-shNSC SYN120 tg littermates). Data are expressed as mean ± SEM. One-way ANOVA + Bonferroni's post-comparison test. n = 6/8 mice for each group.