The Compound ATH434 Prevents Alpha-Synuclein Toxicity in a Murine Model of Multiple System Atrophy

Mice

The MSA transgenic mice used in this study were derived from a colony initially established from breeding animals that were a generous gift from the laboratory of Dr Nadia Stefanova, Clinical Department of Neurology, Innsbruck Medical University, Innsbruck, Austria. These mice express human wild-type α-syn within the oligodendrocytes, driven by the proteolipid protein (PLP) promoter, they show that α-syn is localized only within cells immunopositive for cyclic nucleotide phosphodiesterase oligodendroglia [19]. The phenotype of these mice and pathology has been extensively characterized [20, 21]. The PLP-α-syn mice are maintained on a C57BL/6 background (Animal Resources Centre, Western Australia) and littermate wild types were bred from heterozygous mating at the Florey Neuroscience Institute animal facility for more than 10 generations. Mice were housed in Techniplast brand individually ventilated cages. Mice had ad libitum access to food and water. All procedures involving mice conformed to the Australian National Health and Medical Research Council Code of Practice for the care and use of animals for scientific purposes and were approved by the Florey Institute Animal Ethics Committee. All experiments were designed to minimize the number of animals used, whilst still providing sufficiently powered group sizes to observe effects. Both male (M) and female (F) mice were used in this study. In the 12 MO treated group and control groups –6 F and 3 M were used. In the 16 MO group; Control 9 F 6 M; 3 mg/kg 10 F 5 M; 10 mg/kg 9 F 6 M and 30 mg/kg 9 F 6 M.

Treatment

The mice were treated with either standard, non-irradiated rodent food or the same food but with the addition of ATH434 (provided by Alterity Therapeutics). The compound was incorporated into the food by Specialty Feeds (Glen Forrest, WA) at nominal doses of 0.025 g/kg food (equating to ∼3 mg/kg/day of ATH434 in the mice), 0.083 g/kg food (∼10 mg/kg/day of ATH434 in the mice), and 0.25 g/kg food (∼30 mg/kg/day of ATH434 in the mice). The mice and food were weighed weekly throughout the four months of the experiment. The group that was culled at 12 months of age were treated from 8 months of age; the group culled at 16 months of age were treated from 12 months of age. No mice in any of the treatment groups exhibited any adverse reaction to ATH434. One mouse in the control group was culled for human reasons.

Pole test of motor performance

The pole test was used to assess the speed of a complex motor skill and coordination. In the week before culling, the mice were placed (nose up) on top of a pole (wooden rod; 300 mm long and 5 mm in diameter; attached to a stable weighted base) and the time required for the mouse to rotate 180 degrees (nose down), and to return to the home cage (“Time Complete”) were recorded. The tests were performed on 2 different days (habituation day and test day). Each mouse attempted 5 trials and the fastest time was analyzed [18].

Transcardial perfusion

Following the treatments and the pole motor test was performed, the mice were anaesthetized with 80 mg/kg sodium pentobarbital (Lethabarb, Clifford Hallam, Australia) diluted in 0.9%saline. Mice were perfused and exsanguinated with 50 ml 0.1 M phosphate-buffered saline (PBS). The left-hand side of the brains were post-fixed in 4%paraformaldehyde (Sigma Aldrich, St. Louis, MO, USA) in PBS at 4°C overnight and then immersed in 30%sucrose in 0.1 M PBS at 4°C in preparation for histology. The right-hand side was micro-dissected and stored at –80°C for biochemical analysis and Inductively Coupled Plasma Mass Spectrometry (ICPMS).

Tissue preparation

Substantia nigra samples (containing compacta and reticulata) were homogenized (sonication; Branson, CT, USA) according to their wet weight at 1 : 5 dilution in Dulbecco’s calcium- and magnesium-free PBS (Invitrogen, CA, USA) with a protease in-hibitor (Roche, Basel, Switzerland) and phosphatase inhibitor: phosStop (Sigma-Aldrich). Protein concentrations were determined with a Pierce BCA Protein Assay kit (Thermo Fisher Scientific, MA, USA). One of the duplicate samples was lyophilized for ICPMS analysis, the other frozen for α-syn analysis.

Iron content analysis by inductively coupled plasma mass spectrometry

Samples were weighed, lyophilized, and digested overnight at room temperature in nitric acid (HNO₃) (65%Suprapur, Merck). The samples were further digested by heating at 90°C for 20 min. An equal volume of hydrogen peroxide (H₂O₂) (30%Aristar, BDH) was added to each sample and left for ∼30 min and then heated for 15 min at 70°C. Samples were then further diluted with 1%HNO₃. Metal levels were measured using an Agilent 7700 series ICPMS instrument under routine multi-element operating conditions using a helium reaction gas cell. Results are expressed as microgram of metal per gram of protein weight (μg/g or ng/μg).

Western blot

The tissues were homogenized according to their wet weight at 1 : 5 dilution in Dulbecco’s calcium- and magnesium-free PBS (Invitrogen, #10010049) with a protease inhibitor (Roche, #05056489001) and phosphatase inhibitor: phosStop (Sigma-Aldrich, 4906837001) using a sonicator (Branson, stepped micro-tip). Protein concentrations were determined with a Protein Assay kit (Pierce, #23225). The fractionation procedure commenced with tissue homogenates of a set concentration (125μg in 500μl PBS). These were spun using an ultracentrifuge (Beckman Coulter, California, USA) at 100,000×g for 30 min at 4°C [22]. The supernatant (S1) comprised the soluble fraction; 10μg of protein was loaded on the S1 gels. The remaining pellet was washed 3 times in a PBS buffer and resuspended in 5%SDS (Sigma-Aldrich). The resuspended pellets were homogenized again using sonication and centrifuged at 100,000×g, 30 min at 4°C. The resulting supernatant (S2) was stored and not used. The remaining pellet was washed with 5%SDS/PBS solution and resuspended in 10μl 8%SDS/8M Urea (Sigma-Aldrich) in PBS to solubilize the fraction not soluble in SDS. This solution, soluble fraction 3 (S3), was re-homogenized and incubated at room temperature for 2 h. 5μl of the S3 fractions were reserved for the S3 gels. 10μg of S1 and 5μl S3 fractions were loaded into 4–12%Bis-Tris gels (Invitrogen) with 1X NuPAGE™ MES SDS Running Buffer (Invitrogen). The gel was run at 120V for approximately 90 min and transferred onto PVDF membrane (Invitrogen) using a semi-dry system (Iblot2, Invitrogen).

The membrane for the S1 fraction was stained with 0.1%Ponceau S Solution (Sigma-Aldrich) for at least 5 min to measure total protein levels (using digitized images, LAS-4000 FujiFilm). The stain was removed using Tris-Buffered Saline (TBST pH 8.0) and three washes with MilliQ water. Membranes were then blocked with 5%skim milk in TBST (pH 8.0) for 1 h at room temperature. After a rinse with TBST (pH 8.0), the membranes were incubated with primary antibody (1 : 10,000 LB509 Abcam, Cambridge, UK) followed by rocking overnight at 4°C. The membranes were then rinsed 3 times using TBST and then incubated in secondary antibody (1 : 10,0000, Rabbit-anti-mouse-HRP; Dako, Glostrup, Denmark), rocking for 1 hour at room temperature. Secondary antibody was then washed off and the membrane incubated with chemiluminescence reagents (1 : 1 ratio, ECL, GE Healthcare, Chicago, IL, USA) for at least 1 minute and imaged and analyzed using a LAS-4000 FujiFilm imager and Multi Gauge Software (GE Healthcare).

Immunohistochemistry

Brains were sectioned with a Leica Biosystem Cryostat at –24°C. The left-hand hemisphere was sectioned at 30μm through the SN and Pons, the right-hand hemisphere was used for the biochemistry. The sections were mounted to gelatinated microscope slides (0.5%gelatin in dH₂0) (Grale Scientific cat# S21102A). Sections were stored at –80°C until staining was performed. The slides were removed from –80°C then incubated in a blocking solution (containing 10%normal goat serum in 0.1 M PBS pH 7.4 for 30 min). Slides were then incubated in primary antibody overnight at room temperature (GCI: α-syn: MJFR1, Sapphire bioscience, AB138501 at 1 : 1000). The slides were then washed with PBS three times and incubated in secondary antibody (Goat Anti-Rabbit IgG Antibody, biotin-SP conjugate) for 3 h at room temperature. Slides were then washed with PBS three times and incubated in avidin peroxidase for 1 h at room temperature. The slides were washed again and incubated in 0.05%nickel sulphate and 0.05%cobalt chloride for 20 min and further developed by adding 0.001%H₂O₂ for a maximum of 5 min. Slides were then washed, counterstained with 2%Neutral Red solution for 2 min, dehydrated with increasing concentrations of ethanol (50%, 70%, 90%, and 100%), followed by xylene and mounted with Ultramount No.4 mounting medium (Trajan Scientific, Cat #II 065C). Neutral red is a versatile stain that is used to distinguish neuronal and glial cell membranes, lysosomes, and nuclei.

Stereology of substantia nigra pars compacta

The total number of neurons in the SN was estimated using an optical fractionator sampling design as published previously [23–25] using StereoInvestigator version 11 software (MBF Bioscience, Williston, VT, USA). The regions of interest were traced, and counts were made at regular predetermined grid intervals of x:140μm and y:140μm with an unbiased counting frame of 45μm×45μm and a dissector height of 12μm. Systematic samples of the area occupied by the Neutral Red stained nuclei were made from a random starting point. The average thickness of every section counted was also measured. All stereological analyses were undertaken in a blinded manner, with the data decoded after all the samples had been processed.

Glial cell inclusions stereology

Stereological estimates of GCIs were performed using a fractionator sampling design with the StereoInvestigator version 11 Software (MBF Bioscience, USA). Grid intervals of x: 100μm and y: 100μm with a counting frame of 50×50μm were systemically and randomly positioned by the software using a 63× objective lens (NA 1.63). All stereological analyses were undertaken in a blinded manner, with the data decoded after all of the samples had been processed.

In vitro α-syn assay

To test if ATH434 affected the rate of α-syn disaggregation, in vitro experiments were performed. Purified and lyophilized recombinant human α-syn (Monash Protein Production Unit, Monash University, Australia) was reconstituted with Tris Buffer Saline (TBS, pH 7.4). Pooled aliquots were spun at 100,000×g for 30 min at 4°C to remove pre-formed aggregates/seeds. The supernatant containing the monomeric form was collected and used in the assay. The protein concentration was determined using the BCA method. Iron nitrate was weighed and dissolved in TBS solution. ATH434 was dissolved in 100%DMSO, then diluted to a stock solution using MilliQ water. To each tube, TBS, iron, ATH434/ vehicle and then α-syn was added in sequence with equal concentrations. The final concentration of α-syn, Fe, and ATH434 was 140.0μM. Triplicates of the samples were assayed in the presence of ThT in a Perkin-Elmer Enspire multi-mode plate reader. ThT fluorescence intensity was measured over time at wavelengths of 450 nm emission and 485nm excitation. The relative fluorescent units (RFU) values were normalized to TBS + ThT control wells and were plotted over time. The method has been previously published [11, 18].

Data analysis

The data were plotted as Standard Error of the Mean (SEM). Data were analyzed with either a one-way ANOVA followed by Dunnett’s and Tukey multiple comparisons test, a two-way ANOVA, or unpaired t-tests as appropriate. The software used was GraphPad Prism (GraphPad Software, La Jolla, CA, USA).

Neuroprotective effects of ATH434

At 12 months of age the MSA mice did not show a significant decrease in SNpc neuron number compared to littermate controls, and treatment with ATH434 from 8 months of age did not alter nigral neuronal number (Fig. 1). As such, the rest of the studies focused on older mice. In contrast, the 16-month-old MSA mice had a 37%decrease in the number of nigral neurons compared to controls (thus providing the rationale for our selection of this timeframe as our primary analysis endpoint), and four months of treatment with ATH434 resulted in a significant protection against α-syn mediated nigral cell death (p < 0.001, One way ANOVA with Dunnett’s Post hoc test comparing to the 16-month untreated animals). Each of the doses of ATH434 trialed were equally effective suggesting a threshold effect below 3 mg/kg/day.

OPEN IN VIEWER

Fig. 1

Total Substantia Nigra pars compacta neuron counts in 12-month (Left) and 16-month (Right) PLP-α-syn mice after approximately 4 months of ATH434 or Control diet (Mean±SEM). A) No loss of neurons is observed in the 12-month group. B) At 16 months of age, there was a 37%loss in the number nigral neurons compared to the 12-month-old strain control. Treatment from the age of 12 months with ATH434 reduced the size of this loss to ∼16%at 16 months. One way ANOVA Dunnett’s Post hoc test comparing to the 16-month untreated animals ^***p < 0.001, ^****p < 0.0001. (12 MO WT vs. 16 MO CONT p≤0.0001; CONT vs. 3 mg/kg p = 0.0003; CONT vs. 10 mg/kg p = 0.0012; CONT vs. 30 mg/kg p = 0.0004)

Motor performance

To investigate whether ATH434 altered nigral-mediated complex movement, mice were assessed for their performance on the pole test (Fig. 2A). The time to complete the pole test behavioral task was significantly improved (28%faster) in the MSA mice receiving the ATH434 food, as compared to those provided the control food (ANOVA, F = 3.01, p = 0.04). No dose response relationship was observed.

OPEN IN VIEWER

Fig. 2

A) Time to complete the pole test motor task was improved by treatment with ATH434 in the PLP-α-syn mice after approximately 4 months of ATH434 compared to those eating Control diet (Mean±SEM, Ordinary ANOVA, F = 3.01, p = 0.04). One way ANOVA Dunnett’s Post hoc test comparing to the 16-month untreated animals indicated that the 10 mg/kg reached significance ^*p = 0.02. B) Substantia nigra iron was elevated was elevated in the 16-month PLP-α-syn mice compared to wild type litter mate controls. Four months of treatment with ATH434 in various doses reduced the iron within the SN (Mean±SEM, Ordinary ANOVA, F = 16.03, p < 0.0001). One way ANOVA with Dunnett’s Post hoc test comparing to the 16-month untreated animals indicated that the 30 mg/kg reached significance ^**p = 0.006; WT vs. VEH ^***p≤0.0001

Substantia nigra iron levels

Utilizing inductively coupled plasma mass spectrometry, we demonstrated that iron levels were elevated by 57%in the SN of 16-month-old MSA mice, as compared to littermate wildtype controls. It should be noted that the microdissection included both the substantia nigra reticulata and pars compacta within the dissected tissue block. A primary question in this study was whether this iron elevation was tractable to therapy. As we have previously shown that ATH434 has a dissociation constant (Kd) for Fe(III) and Cu(II) of ∼10^–10 M, and affinities for Fe(II) and Zn of ∼10^–5 M and ∼10^–7 M respectively [18], this is a likely candidate to target iron in the MSA mice. Four months of treatment with ATH434 in various doses reduced the iron within the SN (ANOVA, F = 16.03, p < 0.0001). A Dunnett’s post-hoc test revealed that only the 30 mg/kg treatment group reached significance (p = 0.006, –20%) (Fig. 2B).

Western blot data for α-syn

ATH434 treatment did not alter the levels of soluble monomeric α-syn (14 kDa) but significantly decreased both oligomeric (66 kDa, p < 0.01 by 18%, Fig. 3A–C) and insoluble (urea extracted, p < 0.01 by 69%) α- syn in the SN (Fig. 3D, E). Long term treatment of the mice with the compound appeared to either prevent the formation of higher order species or increased α- syn clearance from the nigra. It should be noted that the microdissection included both the substantia nigra reticulata and pars compacta within the dissected tissue block.

OPEN IN VIEWER

Fig. 3

A) Western blot data for soluble α-syn levels in the Substantia Nigra (SN) in 16-month PLP-α-syn mice after 4 months of ATH434 or Control diet, Soluble (S1) fraction of SN transferred to western blots and stained for α-syn with LB509 antibody. Arrows indicate the α-syn monomer and oligomeric bands. Loading control ponceau stain (P) is show under the labels. B) Plots of density (relative to ponceau) of 14 kDA. C) Plots of density of 36 kDA bands (relative to ponceau), Mean±SEM, T test ^**p = 0.0081. D) Western blot data for aggregated (urea soluble) α-syn levels in the Substantia Nigra in 16-month PLP-α-syn mice after approximately 4 months of ATH434 or Control diet. Urea soluble (S3) fraction of SN transferred to western blots and stained for α-syn with LB509 antibody. The arrow indicates the α-syn urea-soluble band. Laddering of higher order oligomeric bands can be observed in the control 16-month PLP-α-syn mice but are greatly reduced in the animals that have been treated with ATH434. Mean±SEM. T-test ^**p = 0.0049. The S3 plots were constructed using relative fluorescence units.

Glial cell inclusions in the substantia nigra pars compacta

Four months of treatment with ATH434 significantly reduced the number of GCI within the SN (ANOVA, F = 19.98, p < 0.0001). A Dunnett’s post-hoc test demonstrated that each of the treatments significantly reduced the number of GCIs (3 mg/kg (46%p < 0.0001), 10 mg/kg (37%, p < 0.0001), 30 mg/kg (31%, p < 0.0001); Fig. 4).

OPEN IN VIEWER

Fig. 4

The number of Glial Cell Inclusions (dark staining, Arrows) in the Substantia Nigra pars compacta (neurons red staining) in 16-month PLP-α-syn mice. Four months of treatment with ATH434 in various doses reduced the number of GCI within the SN (Mean±SEM, Ordinary ANOVA, F = 19.98, p < 0.0001). One way ANOVA with Dunnett’s Post hoc test comparing to the 16-month untreated animals indicated that 3 mg/kg (46%decrease), 10 mg/kg (37%decrease), 30 mg/kg (31 %decrease) significantly reduced the number of GCI p < 0.0001. (16 MO CONT vs. 3 mg/kg p≤0.0001; 16 MO CONT vs. 10 mg/kg p≤0.0001; 16 MO CONT vs. 30 mg/kg p≤0.0001) Panels show photomicrographs showing GCIs within the SN (scalebar = 25 microns).

In vitro α-syn assay

In an in vitro assay utilizing iron-mediated aggregation of α-syn, we previously demonstrated that the addition of ATH434 slowed the rate of aggregation [18]. An aim for some drug makers is to find compounds that break apart or disaggregate α-syn aggregates. In this study, we used the ThT fluorescence assay to assess if ATH434 could disaggregate already aggregated α-syn. After the day 3 measurements the samples were divided into two groups, control (α-syn + iron + water as vehicle control) or ATH434 (α-syn + iron + ATH434). Visual inspection of Fig. 5 suggests that the average rate of aggregation in the control group is the similar to baseline, whereas the addition of ATH434 the mixture did not allow new aggregates to form and importantly did not disaggregate the already formed fibrils. Analysis of data after treatments were introduced used a two-way ANOVA; samples were paired but the variance component for the pairing was zero. Pair wise multiple comparisons of treatment were performed for each day, as an interaction of treatment and day was detected. Pairwise analysis shows that at day 5 the differences reached a p = 0.013 (^*) and at day 8 p < 0.001 (^**) (Fig. 5).

OPEN IN VIEWER

Fig. 5

Inhibition of α-synuclein (α-syn) aggregation. Recombinant α-syn (140.0μM) was incubated in the presence of equimolar concentrations of iron nitrate (Fe = Fe(NO₃)₃) and or ATH434. The ATH434 was delivered at day 3 after the α-syn aggregation had commenced and prevented further iron-mediated α-syn aggregation. The presence of the ATH434 did not dis-aggregate the α-syn as the signal does not decrease. Aggregation was measured by using ThT fluorescence binding (RFU = relative fluorescent units) p = 0.013 (^*) and at day 8 p < 0.001 (^**)

Mouse colony

Although originating from the same source, the minor drift in the timing of the phenotype observed in the Australian colony may be attributed to different genetic background of the mice or subtle variations in the experimental methods. Heras-Gavin et al. observed loss of TH immunoreactivity in six-month-old animals, consistent with data from Refollo et al. who also found behavioral differences at this time [20, 32]. The Australian colony showed a small but insignificant decrease in total SNpc neuron at 12 months of age, that became significant at 16 months. This finding of temporal differences in the results is not surprising as differences in colonies is known to occur [33]. Importantly GCIs are present in both colonies from an early age.

Alpha synuclein in MSA

It is believed that the pathogenesis of MSA results from spread of misfolded α-syn from neurons to oligodendroglia [34, 35]. The α-syn then accumulates in the oligodendroglia to form toxic fibrils and GCIs. How this accumulation within the glia subsequently causes death of neurons is not clear. Interestingly, it has been shown that the α-syn aggregates that form in the oligodendroglia micro-environment are more toxic and structurally distinct from the aggregates found within the neurons from individuals with PD [3, 36].

In the PLP promoter α-syn transgenic mice, the production of the α-syn is not under the natural promoter that has the iron response element within mRNA. In humans and wild type rodents the iron response element is within mRNA of α-syn [15, 16]. In this study we found that the amount of monomeric α-syn was not altered by ATH434, which is consistent with the α-syn being controlled by the PLP promoter. ATH434 reduced the amount of soluble aggregates and insoluble (urea-soluble) α-syn aggregates that are formed in the oligodendroglial microenvironment. The in vitro study indicated that ATH434 was not able to dis-aggregate α-syn but was able to prevent further aggregation. Therefore, it is deduced that the effects observed in vivo may have been achieved by increasing clearance of α-syn [37, 38] and/or slowing the formation of iron mediated α-syn aggregation. Importantly, this treatment culminated in a reduction in the number of glial cell inclusions which are constituted from aggregated α-syn. Given the data from this manuscript, ATH434 could act on several different pathways to prevent or slow the pathological process. These include potential effects on: 1) levels of synuclein protein, by removing access the iron-response element of the mRNA of α-syn; 2) preventing the formation of extracellular misfolded α-syn; 3) enhancing clearance of the α-syn burden; 4) prevent the formation intracellular α-syn aggregates.

Iron in MSA

Iron accumulates in a specific pattern in MSA[8]. Whilst studies have found that iron deposits in the putamen precede the clinical presentation of MSA by 2 years [39], most find that iron accumulation occurs during atrophy, and this is thought to be a secondary event of the neurodegeneration process [6]. Whilst the precise cellular and subcellular localization of this elevated iron has not been fully established, synchrotron microscopy technology is becoming available to address this question, although it does not yet appear to have been applied to PD or MSA [40].

We have been investigating the normal distribution of iron and the consequences of alterations in the metabolism and homeostasis of iron in parkinsonism [41–45]. We believed that this elevation in iron results in an increase in the weakly bound or labile iron pool [46, 47]. We have shown that this a tractable target and that a reduction in iron is associated with a reduction of damaging radicals which have a number of pathological consequences[18]. In the current MSA study we found that the elevation of iron caused by oligodendroglial α-syn expression could be reduced by the treatment with ATH434 in an apparent dose dependent way.