Husbandry, generation and maintenance of transgenic zebrafish expressing ATXN3

Experiments carried out using animals in this study were performed in accordance with the animal ethics committee of Macquarie University (ARA: 2016/004 and 2017/019). Zebrafish were housed in a standard recirculating aquarium system at 28.5 oC with a 14:10 light:dark cycle and fed daily with artemia and standard pellet. Transgenic MJD zebrafish used in this study had been used and described previously [16]. This transgenic line resulted from crossing driver line zebrafish expressing Tg(elavl3:Gal4-VP16; mCherry) with responder line zebrafish, Tg(UAS:dsRED,EGFP-ATXN3_Q23) or Tg(UAS:dsRED,EGFP-ATXN3_Q84), to generate zebrafish carrying neuronal expression of human ataxin-3 containing either 23Q or 84Q fused to EGFP. Drug treatment studies utilised the resulting offspring (F2) of in-crossing the F1 generation zebrafish.

In experiments that required human ataxin-3 expression without an EGFP-fusion protein, we used a transgenic zebrafish line ubiquitously expressing human ataxin-3 with C-terminal BFP fusion protein. The expression construct was generated using ubiquitin protomer sequence and BFP fusion protein sequence (both from the Zebrafish Tol2 Gateway-Compatible kit (gift from Emily Don, Addgene kit #1,000,000,087, [17]) together with human ATXN3 sequence containing 23 or 84 CAG repeats (pcDNA3-myc-Ataxin3Q84 was a gift from Henry Paulson, Addgene plasmid # 22,124; http://n2t.net/addgene:22124; RRID:Addgene_22124 [18]). These constructs were injected into one-cell stage embryos together with transposase mRNA and embryos carrying BFP (blue) expression were raised to adulthood. Potential founders were outcrossed, and their offspring screened for ubiquitous BFP expression. Tg(-3.5ubb:has.ATXN3-23Q-mTagBFP) and Tg(-3.5ubb:has.ATXN3-84Q-mTagBFP). Offspring from F3 generation of these transgenic lines (named Tg(-3.5ubb:has.ATXN3-23Q-mTagBFP) and Tg(-3u.5ubb:has.ATXN3-84Q-mTagBFP)) were used for the acridine orange staining experiment.

Drug treatment of zebrafish models of MJD

Zebrafish embryos positive for the ATXN3-84Q transgene were identified via the expression of fluorophores (EGFP and dsRED), at 24 h post fertilisation (hpf). Positive embryos were treated with a single administration of spermidine (62.5 µM, 125 µM and 250 µM, solubilised in E3 medium), or chloroquine (1.5 mM and 3 mM, solubilised in E3 medium) and a vehicle control (E3 medium only). Spermidine was purchased from Cayman Chemicals, whilst chloroquine was purchased from Sigma Aldrich. Control (vehicle) treated animals (all genotypes: non-transgenic siblings, EGFP-Ataxin-3 23Q and 84Q) received the equivalent volume of appropriate vehicle (E3 medium). Zebrafish larvae were exposed to the drug compound until 6 days post fertilisation (dpf), at which point motor behaviour analysis and generation of protein lysates for western blotting were performed. Morphologically abnormal larvae were not included in the study. Approximately 20–25 embryos were treated per group per experiment.

Motor behavioural assay of zebrafish larvae

Zebrafish behavioural analysis was performed in the Zebrabox using the Viewpoint Zebralab tracking software. Tracking of 6 dpf larvae was conducted by placement into rows of a 24-well plate, with experimental groups allocated to rows in a randomised manner to eliminate location bias. The multi-well plate was then acclimatised in the Zebrabox for 20 min. Larvae were then exposed to conditions of 6 min light and 4 min darkness. The total distance travelled in periods of darkness were calculated and analysed.

Acridine orange staining of the Ubb ATXN3 BFP zebrafish

Transgenic zebrafish ubiquitously expressing human ataxin-3 (Ubb-ATXN3 23Q BFP and Ubb ATXN3 84Q BFP) as well as wild-type (WTTAB) controls were used for this experiment. Transgenic zebrafish embryos were confirmed positive for the human ATXN3 transgene with the identification of the BFP fluorophore at 24 hpf. Positive embryos, as well as wild-type embryo controls, were treated either with spermidine (250 µM) or vehicle (E3 medium) control from 24 hpf. Zebrafish aged 48 hpf were then treated with acridine orange (diluted into E3 medium; 5 µg/mL) for 10 min before washed in E3 ten times for two minutes each. Zebrafish larvae were individually placed into a 96-multi black well (clear bottom) plate and anesthetised with tricaine. Zebrafish larvae were imaged for GFP using a Leica DMi8 inverted microscope (at 5x magnification) equipped with Leica DFC 365 FX camera (Leica Microsystems). Apoptosis was then measured via GFP fluorescence intensity using Image J [19]. Approximately 15–20 embryos were used per replicate.

CMVMJD135 mouse treatment study

Within this study, we utilised male CMVMJD135 transgenic mice that ubiquitously express human mutant (expanded) ATXN3 driven by a CMV promotor at near endogenous levels [20]. The procedures performed within this study were approved by the Macquarie University Animal Ethics Committee (ARA 2017/044). Tail tissue samples were taken at approximately 2–3 weeks of age and DNA was extracted to confirm presence or absence of the CMV promoter. DNA from animals found to be CMV + then underwent further analysis to sequence the CAG repeat region of the human ATXN3 gene. CMVMJD135 mice with a repeat length of 134 to 141 CAGs were recruited into this study.

The CMVMJD135 transgenic colony was maintained at Australian BioResources (Moss Vale, Australia) and animals were shipped to Macquarie University at 4–5 weeks of age and commenced experimentation at 5.5 weeks of age. Mice were housed with 2–5 mice per cage throughout the duration of the study. A total of 64 male MJD and wildtype littermate mice were recruited into the study, with littermates randomly allocated to either spermidine or water treatment groups by experimenters not involved in the behavioural testing within the study. A total of 36 CMVMJD135 mice (n = 17, spermidine treated and n = 19 water treated) and 27 non-transgenic (n = 12 spermidine treated and n = 15 water treated) littermate controls were included in the study.

Treatment administration of CMVMJD135 mice

Spermidine (Cayman Chemical) was dissolved in acidified drinking water at a concentration of 3 mM for mice whilst control mice received standard acidified drinking water. This dose and administration route was selected based on previous literature demonstrating induction of autophagy in mice following such administration [15, 21]. Mice had ad libitum access to drinking water which was refreshed three times per week (Monday, Wednesday, and Friday). The amount of water consumed by each cage of animals was recorded and compared to identify any difference between the mice administered spermidine versus standard water.

Monitoring of neurological impairment and motor behaviour dysfunction

All mice underwent weekly monitoring for neurological impairment such ashind limb reflex, tremor and ataxic gait through use of a four-point scale (4 indicating highest impairment and 0 indicating no impairment), by a researcher blinded to experimental group. The total score of examined components was calculated for each animal, with twelve representing the highest score possible.

Each mouse was tested on a range of motor tasks including tests for balance and coordination assessment (accelerating rotarod and balance beam). Accelerating rotarod testing measured the latency to fall from an accelerating rotarod (Model 7650, Ugo Basile) with a starting speed of 4 rpm, with gradual acceleration to 40 rpm over the span of 300 s, with a time of 300 recorded for those that did not fall [22]. The beam test measured the latency to cross a square balance beam of 10 mm diameter, performed as described previously [20]. Rotarod and balance beam tests were each performed fortnightly, by a researcher blinded to experimental group, with three repeat tests per test day (each separated by five minutes of rest).

Euthanasia and sample collection

Animals within the treatment study were divided into two cohorts, one cohort was euthanised at 18 weeks old, an age where disease onset is well established and the other at 25 weeks old, an age representing mid-stage disease. Animals were euthanised via sodium pentabarbitone overdose (300 mg/kg, IP) and underwent intracardiac perfusion with 0.9% saline, after which brain and spinal cord tissue was extracted. Brains were hemisected, with one hemisphere snap frozen in liquid nitrogen for protein extraction and the other hemisphere post-fixed in 4% paraformaldehyde for immunohistochemical processing.

Immunohistochemical staining

Following 24 hours in 4% paraformaldehyde, brains were briefly washed in PBS (3 × 2 mins) then placed in 30% sucrose solution for a further 24 hours. Brains were then stored in PBS with 0.02% sodium azide for long-term storage before being embedded in Tissue Tek OCT compound (Sakura Finetek) and cryosectioned at 40 µm before storage in PBS with 0.02% sodium azide at 4°C until use. Collected cryosections underwent free floating DAB (3, 3’ diaminobenzidine) immunohistochemistry for ataxin-3 immunoreactivity. Cryosections were first incubated in 50% ethanol for 20 min at RT, then quenched in 3% H2O2 + 50% ethanol for 30 min at 4 °C prior to blocking in 3% bovine serum albumin (BSA) + 0.25% Triton X-100 in PBS for 1 h at RT. Sections were incubated for 24 h at 4 °C in rabbit anti-MJD (Ataxin-3 1:20,000, gift from Henry Paulson) diluted in blocking solution, followed by overnight incubation at 4 °C in biotinylated secondary antibody (1:2000 Vector Laboratories; goat anti-rabbit BA-1000) diluted in blocking solution. Sections were then incubated for 1 h at RT in Avidin-Biotin complex (Vector Laboratories, Vectastain Elite ABC Kit, Peroxidase Cat#PK-6100), prepared as per manufacturer instructions and left to complex for 30 min before application. Ataxin-3 immunolabelling was detected by incubating in DAB chromogen (Vector Laboratories, DAB Substrate Kit, Peroxidase Cat#SK-4100) for 10 min. Finally, sections were mounted onto glass slides and coverslipped with DAKO mounting media for imaging.

Ataxin-3 imaging and manual quantification of aggregates

Imaging of ataxin-3 DAB sections was performed under brightfield (Zeiss Axio Imager Z3 microscope, 20x objective lens running Zen 3.5 software) at optimised exposure times for 18- and 25-week-old cohorts. Whole section images were captured at 20x magnification using the tiling function, and any stitching errors corrected post-acquisition using Zen lite software (Zeiss, Gottingen, Germany).

Manual blinded quantification of ataxin-3 aggregates within the medulla and pons was performed by a blinded investigator using the Multi-Point counting tool on ImageJ [23]. Required regions of interest were identified using the Allen Brain Atlas [24]. Three anatomically similar sections were chosen per animal for blinded manual aggregate counting within the medulla oblongata and the deep cerebellar nuclei (DCN). The sum of the aggregates from the three sections was taken as the total number of aggregates for the animal. Similarly, due to inconsistent number of hindbrain sections available, aggregates were counted within all anatomically similar sections containing the pons, and the average was taken instead of total aggregate count.

Protein extraction and western blotting

Zebrafish larvae aged 6 dpf were prepared for protein extraction following euthanasia. Zebrafish larvae were placed into RIPA buffer containing protease and phosphatase inhibitors (Complete ULTRA tablets and PhosphoSTOP tablets, respectively; Roche), then homogenised manually using a dounce homogeniser.

Hemisected mouse cerebellar tissue was dissected and homogenised in 5 µL of RIPA buffer containing protease and phosphatase inhibitors per mg of wet tissue weight. Cerebellar tissue was then sonicated using an Omniruptor 250 Ultrasonic Homogeniser (Omni International).

Protein lysates from either zebrafish or mouse brain were centrifuged at 21,300 g at 4 °C for 20 min and clear supernatant collected. Total protein concentration of supernatants was determined using a Pierce BCA Protein Assay Kit (Thermo Fisher Scientific) and equal amounts of protein was prepared in Laemmli buffer (Bio-Rad) with NuPAGE Reducing Agent (Life Technologies). Denatured proteins were separated on NuPage™ 4–12% Bis-Tris gel (Thermo Fisher Scientific) or 4–15% (Biorad) gel through SDS-PAGE, then transferred onto a 0.45 μm PVDF membrane for immunoblot probing.

Antibodies used included rabbit anti-MJD (kind gift from H. Paulson), rabbit anti-phosphorylated ULK1 Ser777 (Merck Millipore), rabbit anti-ULK1 (Cell Signalling Technology), rabbit anti-beclin-1 (Proteintech), rabbit anti-p62 (MBL), mouse anti-p62 (Abcam), rabbit anti-LC3B (Abcam), rabbit anti-PARP (Cell Signalling Technology), rabbit anti-cleaved caspase 3 (Abcam), rabbit anti-caspase 3 (RnD) and mouse anti-GAPDH (Proteintech). Immunoblots were then probed with the appropriate secondary (Promega) and visualised by chemiluminescence (SuperSignal West Femto Maximum Sensitivity Substrate, Thermo Fisher) using either the ImageQuant LAS4000 or ImageQuant 800 Amersham. Band intensity was quantified using Image Studio Lite and the target protein was normalised against the loading control protein (GAPDH).

Statistical analysis

Data analysis was performed using GraphPad Prism (version 9). Group comparisons within the zebrafish studies were analysed using a one-way ANOVA, followed by a Tukey post-hoc analysis. Densitometric analysis of autophagy proteins in the presence of an autophagy inducer versus the vehicle control were compared using a student t-test. Co-treatment studies involving spermidine and chloroquine were analysed using a two-way ANOVA followed by a Tukey post-hoc analysis. The behavioural data within the rodent study was analysed using two-way repeated measure ANOVA with experimental group and age as the factors, followed by Tukey post-hoc comparisons. Comparison of autophagy or apoptosis related protein expression in the mouse study used two-way ANOVA with genotypes and treatment as the effects and Tukey post-hoc comparison. Statistically significant differences are defined as p < 0.05.