Overview of workflow.

[A] Experimental timeline of mouse experiments where wild-type (WT) and M83 hemizygous mice were injected with either phosphate buffered saline (PBS), human (Hu-) or mouse (Ms-) preformed fibrils (PFF) in the right striatum. Imaging and behavioural testing was administered at four time points: −7, 30, 90 and 120 dpi. Mice were siphoned off at each post-injection time point. [B] Neuroanatomical changes over the course of the experiment were assessed to examine MRI-derived PFF-induced atrophy. [C] Motor symptomatology was assessed based on the performance of the mice on the pole test, rotarod and the wire hang test at each of the four time points. [D] Correlative relationship between MRI-derived brain pathology and demographic measures (such as motor symptomatology, injection group, sex, genotype) was examined using partial least squares analysis (PLS). [E] Experimental timeline of the second mouse experiment, such that M83 hemizygous mice received an injection of PBS or Hu-PFF in the right dentate gyrus of the hippocampus to assess aSyn spreading via an alternate disease epicenter. Neuroanatomical changes over time were similarly examined here. [F] Computational modelling of aSyn propagation was generated using the S-I-R model adapted from Rahayel et al. (2022a; 2022b) and Zheng et al. (2019). Accuracy of the in silico model was assessed via comparison of simulated atrophy to empirical atrophy derived from the mouse experiments for both the caudoputamen and hippocampal injection site.

Disease progression and motor symptomatology.

[A] Experimental timeline for WT and M83 hemizygous mice injected with either PBS, Ms-PFF, or Hu-PFF in the right striatum, with MRI and behavioural testing conducted at −7, 30, 90, and 120 dpi. [B] Percent survival plotted across dpi, with the experimental endpoint of ∼130 dpi. M83 PFF-injected mice showed reduced survival, particularly for the M83 Ms-PFF-injected mice (p=5.50e-5). [C] Sex differences in survival rates for M83 Hu-PFF-injected mice, with lower rates in males (p=0.000411). No significant sex differences in survival rates within M83 Ms-PFF injected mice. [D] Weight trend across disease progression showed significant weight loss for M83 Ms-PFF and Hu-PFF-injected mice, with weight loss beginning around 90 dpi (p=0.00119; p=0.00393 respectively) (compared to WT PBS-injected mice). [E] Average rotarod performance across time showed no significant differences between injection groups and genotype across the time points. [F,G] Pole Test performance at 90 dpi [F] and 120 dpi [G]. Pole Test performance at 90 and 120 dpi, indicating worse performance for M83 Ms-PFF mice at both time points (p=0.0499; p=0.000749), and for M83 Hu-PFF at 120 dpi (p=0.00820). [H,I] Wire hang test performance at 90 dpi [H] and 120 dpi [I], performing worse for M83 Ms-PFF at 90 dpi (p=0.000789) and M83 Hu-PFF at both time points (p=0.00691; p=0.000151) compared to WT PBS mice. Purple colour denotes PBS-injected mice, orange colour denotes Hu-PFF-injected mice, and green colour denotes Ms-PFF-injected mice. Line type and shape of points were used to denote each of the genotypes: solid line and round point for wild-type, and dashed line and triangular points for M83 hemizygous mice.

Brain-behavioural correlates of PFF-induced pathology and disease worsening.

[A-E] Coronal slices of the mouse brain (from posterior to anterior) with t-statistical map overlay; demonstrating the effects of [A] M83 Hu-PFF injected mice or [B] M83 Ms-PFF-injected mice (compared to WT PBS mice), [C] M83 Ms-PFF compared to M83 Hu-PFF injected mice, [D] M83 Hu-PFF-injected mice and [E] M83 Ms-PFF-injected mice compared to M83 PBS mice. Colour map denotes significance level of voxel-wise volumetric trajectories over time with cooler colours denoting significant volumetric declines over time while warmer colours denote the opposite for the group of interest relative to the control group. Pink circles denote the peak voxel being plotted. [F-I] Plots of relative volume change (mm3) (compared to the mean volume per group of interest at the pre-injection time point; −7 dpi) over the four time points for a peak voxel in [F] the periaqueductal gray, [G] the right substantia nigra pars compacta (SNc), [H] the injection site (right caudoputamen (CP)), and [I] the right primary motor area (1 MC). Purple colour denotes PBS-injected mice, orange colour denotes Hu-PFF-injected mice, and green colour denotes Ms-PFF-injected mice. Line type and shape of points were used to denote each of the genotypes: solid line and round point for wild-type, and dashed line and triangular points for hemizygous mice. [J,K] Partial least squares (PLS) analysis results for the first latent variable (LV1). [J] Brain loading bootstrap ratios for the LV1 deformation pattern overlaid on the population average, with negative bootstrap ratios in orange-yellow (indicative of negative correlations with behavioural loadings [K]), and positive in blue (indicative of positive correlations with behavioural loadings [K]). Coloured voxels make significant contributions to LV1. [K] Behaviour weights for each behavioural measure included in the analysis showing how much they contribute and the direction of their correlation to the pattern of LV1. Singular value decomposition estimates the size of the bars whereas confidence intervals are estimated by bootstrapping. Bars with error bars that cross the 0 line (blue) are not considered (non-significantly associated). Abbreviations: WH_avg, average wire hang performance; RR_avg, average rotarod performance; PT_avg, average pole test performance.

Simulated spread of aSyn propagation from striatal inoculation, utilizing both aSyn gene expression and structural connectivity, can accurately model empirical atrophy in a mouse model of synucleinopathy.

[A] Spearman correlation between empirical atrophy and connectivity strength (r=0.017) was lower than [B] the correlation between atrophy and SNCA expression (r=0.482). Top correlated atrophy-connectivity regions include the right substantia nigra pars reticulata (SNr) and pars compacta (SNc), the right globus pallidus external (GPe) and internal (GPi) and the right Visceral area (VISC) and the right Agranular insular area posterior part (Alp). [C] Spearman correlations between simulated and empirical atrophy derived from M83 Ms-PFF (the most affected group) compared to WT PBS-injected mice at −7 (orange), 30 (green), 90 (blue) and 120 (pink) dpi. Correlations are shown as a function of simulation time. After reaching the peak value (r = 0.618), the model fit slightly drops and stabilizes. [D,E] Plots of regional pathology (empirical atrophy values at 90 dpi) versus simulated atrophy at the peak model fit (peak correlation time step). Points represent each region of the brain and is colour coded by either [D] its connectivity strength (blue gradient) with respect to the injection site seed, or [E] denotes its local aSyn gene (SNCA) expression (pink gradient) for each region of the AMBA.

Hippocampal seeding of aSyn fibrils results in hippocampal-focused atrophy not well predicted by SNCA local gene expression and structural connectivity.

[A] In vivo T1-weighted MRI (100 μm3 isotropic voxels; Bruker 7T) were acquired at four time points (−7, 30, 90 and 120 dpi) for M83 hemizygous mice injected with PBS or Hu-PFF. [B] Coronal slices of the mouse brain (from posterior to anterior) with t-statistical map overlay; demonstrating the effects of M83 Hu-PFF-induced brain atrophy in terms of voxel-wise volumetric trajectories over time. [C-E] Plots of relative volume change over time (compared to the mean volume per group of interest at the pre-injection time point; −7 dpi) for a peak voxel in [E] the injection site (right dentate gyrus (DG)), and other hippocampal subfields: [C] CA3 and [D] CA1. [F] Spearman correlations between simulated and empirical atrophy derived from M83 Hu-PFF (compared to M83 PBS-injected) mice at −7, 30, 90 and 120 dpi. Correlations are shown as a function of simulation time. After reaching the peak value (r = 0.3297), the model fit slightly drops and stabilizes. [G] Plots of regional pathology (empirical atrophy values at 120 dpi) versus simulated atrophy at the peak model fit (peak correlation timestep); each point represents a region of the Allen Mouse Brain Atlas. [H] Spearman correlation between empirical atrophy at 120 dpi and SNCA expression (r=0.145) and [I] between 120 dpi empirical atrophy and connectivity strength (r=0.167).