Trio^fl/fl mice were purchased from Model Animal Research Center (Nanjing, China). Trio^fl/fl mice were backcrossed with C57BL/6 mice for more than 10 generations. Ai14 mice (Express tdTomato fluorescence following Cre-mediated recombination, Jax007914) and Pcp2-Cre (express Cre recombinase under the control of the mouse PCs protein, Jax004146.) mice were supplied by the CIBR from Jackson lab. We generated Trio^{fl/fl;Pcp2-Cre} mice by crossing Trio^fl/fl mice with Pcp2-Cre mice. While Trio^{fl/fl; Pcp2-Cre ;} Ai14 mice was generated by Trio^fl/+;Ai14 and Trio^{fl/+;Pcp-2Cre}. Animals were kept in standard environmental conditions with temperatures maintained between 22-25°C and a 12h light/12h dark cycle. All animal care and handling procedures were strictly in compliance with the ethical committee’s regulation for use and management of experimental animals at the CIBR (LARC-T019, CIBR-IACUC-023). The design and conduct of the experiments made every effort to adhere to the principles of animal welfare. At least 24 hours prior to behavioral experiments, mice were placed in the testing room to acclimate to the behavioral testing environment, which was designed to reflect their husbandry conditions. In addition, the mice were handled to become familiar with the experimenters. For MRI studies, mice were acclimated to the experimental environment 24 hours in advance. Following the completion of experiments, mice were euthanized by cervical dislocation under 1.25% tribromoethanol anesthesia or by perfusion with cold PBS for tissue collection.

To assess overall changes in the cerebellum of mice, we weighed the cerebella of knockout and control mice with similar body weights (25-30g). Using blunted dissection, intact cerebella were harvested, weighed, and the size was measured across two orthogonal axes. For further quantification of cerebellar size, we calculated the cross-sectional area of cerebellar slices taken along the vermis’s mid-sagittal plane. The mice were anesthetized and perfused with cold PBS and 4% paraformaldehyde. The brains were dehydrated with gradient sucrose, and the frozen sections were prepared with a thickness of 30μm, and photographed under fluorescence microscopy. Measure total cerebral area, molecular layer (ML), granular cell layer(GCL),and cerebral white matter(WM) area. Areas of all layers were computed using Imaris software. Also quantified was the thickness of the molecular layer across the first four lobules of the vermis in median sections. These data were derived from at least three pairs of mice, with a minimum of five slices per mouse.

The mice were anesthetized with 5% sodium pentobarbital, and perfused with pre-cooling 4°C PBS. The cerebella were extracted and lysed with RIPA lysis buffer by using ultrasonic disintegration on ice. The protein concentrations were determined by using a BCA protein assay kit. Proteins were separated by electrophoresis in a gel using 80mV for 2 hours and transferred onto NC membranes at 220 mA for 1.5 hours. After blocking with 5% non-fat milk, the membranes were incubated with primary antibodies overnight at 4°C, followed by one hour with secondary antibodies at room temperature. The antibodies used in the experiment are listed below: Monoclonal Anti-Calbindin-D-28K antibody (C9848,1:20000, Millipore Sigma); Rabbit monoclonal to Syne-1(ab192234, 1:1000,abcam, Shanghai), Anti-GAPDH antibody (2118,1:2000, Cell Signaling Technology), Goat Anti-Mouse IgG H&L (HRP)(IH-0011, 1:2000, Dingguo), Goat Anti-Rabbit IgG H&L (HRP)(IH-0031, 1:2000, Dingguo). Images were captured using an ECL system within a gel imager. Data analysis included calculating the intensity of bands using Image J software.

For MRI scans, the mice were fixed on an animal bed (Bruker, Ettlingen, Germany), with the head fixed using dental strips and dental pads. The mice were anesthetized by c1.25% tribromoethanol anesthesia intraperitoneal injection. The respiratory and heart rates as well as the body temperature of the mice were recorded to maintain a body temperature of 37.0 ± 0.5°C. Blood oxygenation was detected using a fiber-optic pulse oximetry sensor located on the foot of the mice. At the end of the MRI session, the mice were sacrificed under anesthesia by cervical dislocation.

A Bruker Pharmascan 9.4 T imaging system (Bruker, Ettlingen,Germany) was used. In each animal, the following sequences were used: 1) two-dimensional Turbo Rapid Imaging with Refocused Echoes T2-weighted images (in three orthogonal directions) while positioning with the following scanning parameters: repetition time 2600 ms; echo time 33ms; field of view 18 × 15 mm²; pixel size 0.07 × 0.059 mm²) two-dimensional DTI scans using echo-planar imaging sequences with the scan parameters: repetition time 7,000 ms; echo time 33 ms; 22 coronal slices 0.7 mm; slice gap 0.07 mm; b-value 1,000 s/mm²; diffusion gradient pulse duration δ 4 ms. Left and right frequency codes were used to extend the diffusion sensitization gradient in 45 optimal directions with a flip angle of 90°. Fifteen b0 images (b = 0 s/mm² and 5 b0 images per 20 diffusion-weighted images) were acquired. The total acquisition time was approximately 20 min.

The statistical parametric mapping-compatible DTI toolbox was used in the statistical process for data calibration and voxel-based analysis. The DTI Studio software1 was used to compensate for motion, and the diffusion-weighted image was linearly registered. DTI Studio calculated the Apparent Diffusion Coefficient (ADC) values. All normalized ADC images were cropped to 1.0 × 1.0 × 1.5 mm³ voxels (after zooming) for Gaussian smoothing (2.0 × 2.0 × 4.0 mm³ voxels [Gaussian kernel]), and a voxel-based analysis was performed. In the entire brain, 100 + voxel clusters with significant differences in ADC between groups (p < 0.005) were marked. The statistical analysis was performed using the SPM12 and spm rat IHEP toolboxes2. The data were spatially normalized to allow for voxel comparisons between animals. Data were compared between groups using an analysis of variance (ANOVA). The critical t-value was determined with 3.169 [t(10) = 3.169, p = 0.005].

The Cerebellum from 6 mice aged 12week were isolated and harvested in RNA later. Cerebellum lysates from 3 mice in each group were collected for total RNA isolation, following the instructions of the manufacturers of the Trizol and RNeasy Lipid Tissue Mini Kit (QIAGEN, Hilden, Germany). Samples were eluted in 15μL of RNase-free H2O and quantified using a Nanodrop ND-1000 (Thermo Fisher Scientific, Inc., Waltham, MA, USA) spectrophotometer. Agilent 2100 Bioanalyzer and 2100 RNA nano 6000 assay kit (Agilent Technologies) were used to evaluate the integrity of RNA samples. All samples were adjusted to 100 ng/μL and only used when the RNA integrity number (RIN) was ≥9.0. Library preparation and sequencing were performed with NextSeq500 (Illumina, Inc., San Diego, CA, USA) at Capital Bio Technology (Beijing, China). After sequencing, the color-space data was demultiplexed and aligned against the Mus_musculus.GRCm39/mm10 mouse genome with LifeScope using the paired-end whole transcriptome module (Thermo Fisher Scientific, Waltham, MA, USA).

Raw data (raw reads) of fastq format were firstly processed through in-house perl scripts. In this step, clean data (clean reads) were obtained by removing reads containing adapter and trimming low quality base with Trimmomatic. At the same time, Q20, Q30 and GC content the clean data were calculated. All the downstream analyses were based on the clean data with high quality.

(For DESeq2 with biological replicates) Differential expression analysis of two conditions/groups (two biological replicates per condition) was performed using the DESeq2 R package (1.16.1). DESeq2 provide statistical routines for determining differential expression in digital gene expression data using a model based on the negative binomial distribution. The resulting P-values were adjusted using the Benjamini and Hochberg’s approach for controlling the false discovery rate. Genes with an adjusted P-value <0.05 found by DESeq2 were assigned as differentially expressed.

(For edgeR without biological replicates) Prior to differential gene expression analysis, for each sequenced library, the read counts were adjusted by edgeR program package through one scaling normalized factor. Differential expression analysis of two conditions was performed using the edgeR R package (3.18.1). The P values were adjusted using the Benjamini & Hochberg method. Corrected P-value of 0.05 and absolute foldchange of 2 were set as the threshold for significantly differential expression.

To confirm the results of sequencing, firstly, the simple random sampling method was used in selecting RNAs with random numbers generated in Microsoft EXCEL. Total RNAs of cerebellum in mice were extracted using TRIzol reagent (Invitrogen, USA), and reversely transcribed by reverse transcriptase according to the manufacturer’s protocol. Then, RT-qPCR was performed with SYBR Green qPCR Master Mix (Bio-Rad, USA) on CFX96 real-time PCR system. The used primers were listed in Table 1 . The relative quantification of mRNAs were normalized to GAPDH with the 2^−ΔΔCT method.

The resulting counts tables were loaded into R (version 3.5.1). Counts data were transformed using DESeq2’s rlog function. Surrogate variable analysis (sva, v3.28.0) was used to remove unwanted variation based on the study design and the detected surrogate variables were regressed out of the normalized count matrix. The normalized, SVA-corrected data were used in WGCNA analysis using the block wise Modules function. For each gene, we report the kME (correlation of a gene with the eigengene of the module) across all modules, the p-value across all modules, the module assigned by WGCNA, and whether the WGCNA-assigned module had the best p-value or kME for that module.

Data are expressed as mean ± Standard Error of the Mean (SEM) and analyzed using GraphPad Prism 8 software (La Jolla, California, USA). Data followed normal distribution as evidenced by the Shapiro–Wilks normality test with the hypothesis for normality rejected at a p value less or equal to 0.05. and for homogeneity of variance using the Levene’s test. Data were analyzed using Student’s t test. For all analyses Statistical significance was set at *p < 0.05, **p < 0.01, ***p < 0.005.

The spontaneous locomotion activity of Trio^{fl/fl;Pcp2-Cre} (cKO) mice was assessed by using the open field test at both 12-week-old and 20-week-old, with Trio^fl/fl mice as the control group. Compared to their counterparts, the cKO mice exhibited a significant decrease in total distance traveled over a 2-hour period ( Figure 1A ), at both ages evaluated. For evaluating the motor coordination, the rotarod test was employed. No significant differences in latency to fall or falling speed were observed between the cKO and control groups at the age of 12 weeks ( Figure 1B ). However, at 20 weeks of age, the cKO mice demonstrated a significant reduction in latency to fall and falling speed on Day3 ( Figure 1C ). The balance beam test was performed to evaluate the balance ability and fine motor coordination. Slight increases, but statistically insignificant, was observed in the cKO mice at the age of 12 weeks compared to control, with no substantial increase in the number of foot slips ( Figure 1D ). In contrast, a significant increase in the number of foot slips was noted in the 20-week-old cKO mice ( Figure 1D ), suggesting impaired motor function in the cKO mice at this age. These results indicated that the mice with Trio deletion in PCs displayed progressive motor impairments.

We performed gait test analysis to evaluate the gait characteristics of cKO mice ( Figure 2A ). The results showed that the decline of stride length of both the front and rear limbs were statistically significant between 20-week old WT and cKO mice (p < 0.05) ( Figure 2B ). The front and rear limb distances and the overlap of footprints did not show significant differences between the two groups ( Figure 2C ). However, no significant differences were observed between the two genotypes in the 12-week-old mice ( Figure 2D ).

We first determined the size of cerebellum in cKO mice and their control littermates in two principal orthogonal axes. There were no significant differences between the two genotypes in the 12-week-old mice ( Figure 3A ). Meanwhile, the weight of cerebellum in two group is not significant ( Figure 3B ). We measured the total area of the mid-sagittal cerebellum and found a significant decrease in cKO mice compared to the controls ( Figure 3C , p < 0.05) in 12-week-old mice. Further analysis revealed that the reduction of the molecular layer (ML) and white matter (WM) area in cKO mice, with statistically significant differences ( Figure 3C , p < 0.05). We also measured the averaged ML thickness in the mid-sagittal sections of cerebellar vermis lobes I/II, III, IV/V, and VI and found varying degrees of reduction in different lobes in the 12-week-old mice. Lobes I/II (p < 0.001) and IV/V (p < 0.001) showed statistically significant differences. Lobe III and VI showed a trend of decreased thickness, but the differences were not statistically significant in the 12-week-old mice. Additionally, these findings indicate that significant alterations in the cerebellum occur at 12 weeks of age, affecting major lobes related to motor function.

Then we determined the morphological change of cerebellum in 20-week-old mice with Trio deletion in PCs. We measured the length of cerebellum in cKO mice and their control littermates in two principal orthogonal axes. There were no significant differences between cKO mice and controls in 20 weeks ( Figure 4A ). The weight of cerebellum also showed no significant difference between the two genotypes ( Figure 4B ). The total area of the mid-sagittal cerebellum show a significant decrease in cKO mice compared to controls ( Figure 4C , p < 0.05) in the 20-week-old mice. Further analysis revealed that the reduction in the ML and WM area, with statistically significant differences ( Figure 4C , p < 0.05) in the 20-week-old mice. We also measured the average ML thickness in the mid-sagittal sections of cerebellar vermis lobes I/II (p < 0.001), III (p < 0.05), IV/V (p < 0.05), and VI (p < 0.05) and found varying degrees of reduction in different lobes in the 20-week-old mice, with statistically significant differences. In mice aged 20 weeks, there was an increased number of affected lobules compared to 12-week-old mice, indicating that the Trio deletion-induced impairment was more severe in the 20-week-old mice.

Furthermore, the expression levels of the PCs’ marker Calbindin was persistently decreased in both 12-week-old and 20-week-old mice ( Figures 3D , 4D ), with statistically significant differences (p < 0.05).

To further investigate the white matter of the brain changes in the cerebellum, we performed DTI-MRI measurements. T2 and DTI sequences were acquired. The VBA algorithm revealed a significant decrease in apparent ADC ( Figures 5A, B ) values in the first cerebellar lobule of 12-week-old cKO mice compared to the control group (p < 0.05).

To gain further insights into the difference of downstream gene expression levels in the cerebellum by Trio deletion in PCs, we performed RNA sequencing. We identified 158 upregulated genes and 75 downregulated genes in cKO group compared to the control group ( Figure 6A ). Differentially expressed gene clustering analysis revealed the involvement of multiple pathways, including cell cytoskeleton, cell apoptosis, and neurodegenerative diseases ( Figure 6B ). Among the 158 upregulated genes, we found that Syne1 was associated with cerebellar ataxia ( Figure 6C ). Co-expression analysis of Syne1 revealed co-expressed genes, including Hs6st3, Tnfrsf18, Stxbp5l, Soga1, Hs6st3, and Lin7a, with statistically significant differences ( Figures 6D–F ). We performed verification of mRNA expression profiles with RT–qPCR. The expression level of mRNA and protein in Syne1 was decreased in 12-week old Trio^{fl/fl;Pcp2-Cre} mice compared with controls, with statistically significant differences ( Figures 6G, H ).