Restoring calcium homeostasis in Purkinje cells arrests neurodegeneration and neuroinflammation in the ARSACS mouse model

Autosomal recessive spastic ataxia of Charlevoix-Saguenay (ARSACS) is caused by mutations in SACS gene encoding sacsin, a huge protein highly expressed in cerebellar Purkinje cells (PCs). Patients with ARSACS, as well as mouse models, display early degeneration of PCs, but the underlying mechanisms remain unexplored, with no available treatments. In this work, we demonstrated aberrant calcium (Ca2+) homeostasis and its impact on PC degeneration in ARSACS. Mechanistically, we found pathological elevation in Ca2+-evoked responses in Sacs–/– PCs as the result of defective mitochondria and ER trafficking to distal dendrites and strong downregulation of key Ca2+ buffer proteins. Alteration of cytoskeletal linkers, which we identified as specific sacsin interactors, likely account for faulty organellar trafficking in Sacs–/– cerebellum. Based on this pathogenetic cascade, we treated Sacs–/– mice with Ceftriaxone, a repurposed drug that exerts neuroprotection by limiting neuronal glutamatergic stimulation and, thus, Ca2+ fluxes into PCs. Ceftriaxone treatment significantly improved motor performances of Sacs–/– mice, at both pre- and postsymptomatic stages. We correlated this effect to restored Ca2+ homeostasis, which arrests PC degeneration and attenuates secondary neuroinflammation. These findings disclose key steps in ARSACS pathogenesis and support further optimization of Ceftriaxone in preclinical and clinical settings for the treatment of patients with ARSACS.

Immunofluorescence of sacsin and its interactors was performed in primary PCs as described above, with the addition of following steps for sacsin detection: after cell fixation, antigen retrieval was performed incubating cells in Tris 10mM, EDTA 1mM, Tween-20 0.05% at pH 9.0 for 30' at 70-90°C; then cells were washed with DPBS1X and pre-blocked for 20' with 50mM glycine in DPBS1X; Bovine Serum Albumin 0.2% was added in blocking buffer.

Synaptosome purification from mouse cerebellum
Mouse cerebellum was homogenized in sucrose buffer (0,32 M sucrose, 4 mM HEPES pH 7.4 and protease inhibitors). Samples were then centrifuged at 1000 g 10' at 4°C to pellet debris and nuclei. Supernatant was centrifuged at 10000 g 15' at 4°C to obtain crude synaptosomal fractions. The pellets were washed once in the same buffer and centrifuged again at the same speed. Pellets underwent an osmotic shock by resuspending in aqueous buffer (4 mM HEPES pH 7.4 and protease inhibitors) and rotating 30' at 4°C. They were then centrifuged at 25000 g 20' at 4°C obtaining the synaptosomes.

EM image analysis
EM were done in collaboration with the Unit of Neuropathology of San Raffaele Institute.
Cerebellum was incubated in glutaraldehyde fixative to perform standard EM experiments.
Several cells were evaluated in a single sample for a total 3 Sacs -/versus 3 wild-type tissues.
Morphometric analysis of mitochondria in EM images was performed using the freehand line selection tool of ImageJ software to calculate mitochondria area and perimeter, major and minor axis, circularity, and roundness. Circularity is 4π*area/perimeter^2. A value of 1.0 indicates a perfect circle. While roundness is the inverse of the aspect ratio (major axis/minor axis) and is calculated as follows: 4 * area/(π * major_axis^2).
To assess the health state of mice treated with Ceftriaxone, just before the sacrifice, as much blood as possible was collected via eyeball. The blood was then analysed to measure haematological values and clinical chemistry parameters to evaluate kidney and liver functionality by following standard mouse clinic protocols.

Supplemental figure 1. (A) WB analysis showing levels of npNFH (lower band as indicated by the arrow) in
Sacs -/and wild-type control cerebellum at P15 and 1 month of age with relative quantitation (normalized to calnexin, which was not significantly altered in LFQ proteomics, see Supplemental Table 2 Pre-treatment (5 month of age) motor assessment by BW test readouts of latency time to cross the beam and number of hindfoot missteps both in female and male mice. Bars represent mean±SEM; n= at least 10; Welch's t-test: **p<0,01, ****p<0,0001; (B) BW test performance in term of latency time to cross the beam and number of hindfoot missteps at 7 months of age male Ceftriaxone-and vehicle-treated mice. Bars represent mean±SEM; n=at least 6; Two-way ANOVA with Tukey's correction: ****p<0,0001; (C-D) WB analysis showing levels of IP3R1 and Calbindin (C) and Parvalbumin (D) in Sacs -/vehicle-and Ceftrixone-treated cerebellum at 7 month of age (post-symptomatic Ceftriaxone treatment) with relative quantitation (normalized to calnexin). Bars represent mean±SEM; n=at least 3; Welch's t-test: *p<0,05; (E) Mus musculus NF-kappa-B putative binding sites (p-value <0.001). Data of putative binding sites (red hits) were retrieved for Nfkb1 and Nfkb2 with EPD search motif tool (Eukaryotic Promoter Database, available at https://epd.epfl.ch/), scanning from -2000 to +100 in respect to TSS, according to JASPAR core 2018 vertebrates library. Nfkb1 and Nfkb2 encode respectively for the p105 and p100 DNA-binding subunits of NF-kappa-B transcription factor.