Gba+/+;Gbatg and Gba−/−;Gbatg mice, which contain a Gba transgene regulated by doxycycline, were generated and maintained as described [11], with two modifications (to enhance breeding efficiency) inasmuch as doxycycline was added to the drinking water of parental cages, rather than at six weeks after birth, and at a lower concentration (10−7 mg/ml). This affected the survival time of the mice, which now lived for 110–120 days after birth, with symptoms beginning to appear by ~60 days of age. Following breeding and stereotaxic injections (described below), four groups were used: Gba+/+;Gbatg injected with PBS, n = 10; Gba−/−;Gbatg injected with PBS, n = 10; Gba−/−;Gbatg injected with AAV-wtGCase, n = 8; Gba−/−;Gbatg injected with AAV-dGCase, n = 12. Group size was determined using the ‘pwr’ package in R [one-way ANOVA (analysis of variance) test for a general statistical power of 0.7, medium effect size of 0.5 and statistical significance level of 0.05], justifying a sample size of 10 mice per group. Procedures involving animals were reviewed and approved by the Institutional Animal Care and Use Committee of the Weizmann Institute of Science (IACUC application number 05310622-2), following guidelines issued by the Association for Assessment and Accreditation of Laboratory Animal Care. No randomization or blinding were performed.

Male and female Gba+/+;Gbatg and Gba−/−;Gbatg mice were genotyped at 20–25 days of age, about a week prior to weaning. DNA was extracted [13] and PCR performed in a QuantStudio™ 5 (Applied Biosystems, Singapore); initial denaturation 95 °C, 3 min; 10 touchdown cycles with 15 s of denaturation at 95 °C, 15 s of annealing from 65 °C to 60 °C, and 10 s of extension at 68 °C; 28 cycles with 15 s of denaturation at 95 °C, 15 s of annealing at 60 °C, and 10 s of extension at 72 °C; and a final stage at 60 °C to 98 °C at a ramp mode set to 0.025 °C/sec (see Supplementary Table 1 for the primers used). Mice which survived until ~120 days of age were genotyped again for validation.

Vectors were generated at the translational vector core of the University Hospital of Nantes by packaging AAV2-based recombinant genomes containing DNA sequences encoding wtGCase or dGCase, under the control of a ubiquitous CAG promoter, into AAVrh10 capsids using helper virus-free transfection of HEK293 cells [8]. The vectors were purified using an optimized CsCl gradient-based purification protocol [14]. Viral protein purity and identity were verified by SDS-PAGE and silver staining, and vector titers quantified by qPCR with primers targeting the flanking sequence of ITR2.

The method for stereotaxic injections was modified from ref. [15]. 35–38-day-old mice were injected intraventricularly into the left ventricle with 4 µl of AAV-wtGCase or AAV-dGCase [both at 1.1 × 1013 viral genome(vg)/ml] at a rate of 0.3 µl/min using a Hamilton syringe (10 µl with 26 G straight Hamilton needle, 22 mm PT3) to coordinates relative to the Bregma anterior-posterior (0.57 mm), medial-lateral (1.17 mm), dorsal-ventral (2.04 mm). Mice received painkillers for 2 days post-surgery (Carprofen®) and were monitored daily for distress, pain and discomfort. Mice were weighed and examined every day. Mice were humanely sacrificed if motor symptoms deteriorated, if they showed any other signs of distress or discomfort, or if they lost 20% of their body weight.

Homogenates of brain tissue were prepared as described [11]. Cell pellets were lysed in ultra-pure water containing a protease inhibitor cocktail (Sigma; 1:100) and homogenized using a GentleMACS dissociator (Miltenyi Biotec). Protein was measured by the BCA method prior to lyophilization. Quantitative analysis of GlcCer in lyophilates (600 µg protein/ml) was performed by liquid chromatography-tandem mass spectrometry [16].

Enzymatic activity was determined using a fluorescently labeled substrate of GCase, C6-NBD-GlcCer [17, 18]. The assay was performed using homogenates (10 µg of protein) from brain tissue in a final volume of 20 µl McIlvaine buffer, pH 4.2. The reaction was run at 37 °C for 60 min and terminated by addition of 1.5 ml of chloroform-methanol (1:2, v/v) prior to lipid extraction.

Motor function assessment was carried out in the middle of the dark-active phase of the diurnal cycle in designated experimental rooms. In order to minimize stress, mice were kept in their home cages for 1 h in habituation locker-like cabinets. Mouse coordination and balance were assessed by RotaRod (San-Diego Instruments) and beam walk tests.

Rotarod sessions consisted of 5–6 trials with an inter-trial interval of 1 min (adapted from Ref. [19]). In each trial, mice were placed on the rod while it was motionless and the rod accelerated the rotation speed from 0 to 40 r.p.m. over 4 min. Mice latency to fall off the rod was recorded and a mean score was calculated for each mouse based on the 3 best performances from all trials.

In the beam walk test [20], mice were first trained to walk on a plank (50 cm long and 35 mm wide), suspended 30 cm above the working surface. Following this, mouse balance and coordination were assessed five times while walking on the plank without stops or turns. These runs were recorded using an overhead camera and manually analyzed for the number of steps for each paw as well as the number of slips. The percent of slips per step was calculated. If mice were dragging their bodies or grasping the plank on the side, the run was designated as 100% slip.

MRI experiments were performed on a 9.4 Tesla BioSpec Magnet 94/20 USR system (Bruker, Germany) equipped with a gradient coil system capable of producing a pulse gradient of up to 40 gauss/cm in each of the three directions. All MR images were acquired with a quadrature mouse head surface coil and transmitter linear coil (Bruker). The MRI protocol included high resolution multi-slice T2 weighted images and T2 maps. The T2 weighted images were acquired using the Rapid Imaging with Refocused Echoes (RARE) imaging sequence with the following parameters: a repetition delay of 1800 ms, effective time echo of 25 ms, RARE factor of 8, matrix dimension of 180 × 180 and 8 averages, corresponding to an image acquisition time of 5 min 28 s. Nineteen continuous slices with a slice thickness of 0.50 mm were acquired with a field of view of 1.8 × 1.8 cm2.

T2 maps were acquired using the multi-slice spin-echo imaging sequence with the following parameters: a repetition delay of 3000 msec, 16 time echo increments (linearly from 10 to 160 msec), matrix dimension of 256 × 128 (interpolated to 256 × 256) and 2 averages, corresponding to an image acquisition time of 12 min 48 s. The T2 dataset consisted of 16 images per slice. Thirteen continuous slices with a slice thickness of 0.90 mm were acquired with a field of view of 2 × 2 cm2. To evaluate volumetric changes between mouse groups, the whole brain and cerellbelum were segmented using the high-resolution T2-weighted images using Bruker ParaVision 360 software.

A quantitative T2 map was produced from the multi-echo T2-weighted images. The multi-echo signal was fitted to a mono-exponential decay to extract the T2 value for each image pixel. Image analysis was performed using homemade scripts written in Matlab R2013B. The procedure of co-registration (inter-subject and intra-subject) was applied before the MRI dataset analysis. For optimal suitability to a mouse brain atlas (correction of head movement image artifacts), all images went through atlas registration: reslicing, realignment and smoothing, using SPM software (version 12, UCL, London, UK). Voxel-by-voxel one-way analysis of variance [ANOVA; (p < 0.05)] comparison of the mice groups was done by SPM software. An effect was considered significant at p < 0.05. One-way ANOVA corrected for multiple comparisons with the false discovery rate approach (p < 0.05 following correction).

Total RNA was isolated from three brain areas using a RNeasy mini kit (Qiagen, Germany) according to manufacturer’s instructions. cDNA synthesis was performed using a High-Capacity cDNA Reverse Transcription kit (Applied Biosystems, USA). Quantitative PCR was performed using the Fast SYBR™ Green Master Mix (Applied Biosystems, USA) and ABI Prism 7000 Sequence Detection System (Applied Biosystems, USA) with cDNA with equivalent of 5 ng of total RNA. The primer concentration was 100 nM in a reaction volume of 10 µl. The thermal cycling parameters were as follows: step 1, 95 °C for 10 min; step 2, 95 °C for 15 s; step 3, 60 °C for 30 s; step 4, 68 °C for 30 s. Step 2 was repeated for 40 cycles and was followed by a dissociation step. Primers are listed in Supplementary Table 2.

Statistical analysis was performed using R studio (version 2023.06.1 + 524). Data modeling was performed using “lme4” package (mixed-effect models) and “survival” package (Kaplan-Meier curve). Differences in mean values between groups were tested using ANOVA followed by Tukey’s Honest Significant Difference (HSD) post-hoc test. A mixed ANOVA model was applied in cases of 2 or more measured effects (using the ‘aov’ function). Assumptions of normality (Q-Q plot and Shapiro’s test) and equality of variances were tested (Levene’s test, package “car”). Graphs were generated using “ggplot2” package.