Prion Protein Deficiency Results in Synaptic, Neural Network and Behavioral Alterations

Key results

1. ShownBoth knockout lines lose excitatory and presynaptic release proteins. PSD95, VGLUT1, SYT1 and CPLX1-2 fall in mature FVB knockout (koFVB) cultures and in adult cortex of co-isogenic ZH3 knockouts. VGLUT2 is spared, and GAD1/GAD2 drop only in koFVB.

   Western blots normalised to β-actin, unpaired two-tailed t-tests. koFVB cultures at 23 days in vitro (DIV23): PSD95 p<0.0001, VGLUT1 p = 0.0022, SYT1 p = 0.0132, CPLX1-2 p = 0.0050, GAD1 p<0.0001, GAD2 p = 0.0031, VGLUT2 p = 0.2209; most markers unchanged at DIV10. 3-month ZH3 cortex: PSD95 p = 0.0065, VGLUT1 p = 0.0005, SYT1 p = 0.0092, CPLX1-2 p = 0.0003; VGLUT2, GAD1, GAD2 not significant. · Synaptic pathways are disrupted in PrPC…; Conserved glutamatergic and synaptic-marker…; Figure legends

2. SuggestedMature (DIV23) koFVB cultures fire as many spikes as wild type but have rarer network-wide bursts and higher pairwise rate correlation. Burst rate and correlation do not differ at DIV10 or DIV17, though burst amplitude was lower in koFVB at DIV17 (adjusted p = 0.0500). ZH3 cultures show no significant difference.

   30-min recordings on 120-electrode MEAs; clustered Wilcoxon with Holm adjustment. DIV23 inter-burst interval (IBI; median of MEA medians) 3395 vs 9518 ms, wtFVB vs koFVB (6 vs 6 MEAs; adjusted p = 0.0260, rank-biserial r = −0.635); cross-correlation index 0.09097 vs 0.4480 (5 vs 5 MEAs; adjusted p = 0.04762). DIV23 total spikes p = 0.3654. ZH3 (1 preparation): IBI adjusted p = 0.5893 to 0.6000. · PrPC Deletion Impairs Synaptic Network Activity…; Conserved glutamatergic and synaptic-marker…; Figure legends

3. SuggestedAdult ZH3 mice match wild type in anxiety, exploration and sociability assays but escape more often from a dimming disk, a stimulus that rarely triggers escape in wild type. They also keep escaping across repeated looms.

   Dimming disk: escape 63% vs 27% (ZH3 vs WT; binomial mixed model β = 1.62, SE = 0.78, p = 0.04; 14 vs 8 mice). Looming: latency 0.94±0.15 vs 1.56±0.37 s (p = 0.058), escape 67% vs 50% (p = 0.52). Open field, dark-light, plus maze, novel object, marble burying and social tests: no significant genotype effect (14 vs 8 mice for open field, 4 vs 5 otherwise). 20 repeated looms: 5 ZH3 mice, no WT group reported. · Behavioral flexibility but not states are…; Behavioral and neuronal responses to danger…; Figure legends

4. SuggestedIn head-fixed adult ZH3 mice, Neuropixels recordings show stronger superior colliculus (SC) responses to looming and dimming stimuli, fewer spontaneous bursts in the periaqueductal gray and higher correlation among cortical neurons.

   Hierarchical Bayesian models compared by cross-validated predictive density: grouping neurons by genotype gave a robust looming difference in deep but not superficial SC, and grouping by mouse added no predictive power (n = 6 per genotype). Periaqueductal gray burst rate 0.0009 vs 0.0006 bursts/s (WT vs ZH3, 89 vs 129 cells, p = 0.034); cortex p = 0.442; correlation shown without a statistic. · Differences in spontaneous neural activity also…; Behavioral and neuronal responses to danger…; In vivo data model; Figure legends

5. ShownThe two lines' transcriptomes diverge at DIV23: koFVB cultures have a small set of differentially expressed cytoskeletal and synaptic genes, ZH3 cultures almost none. The authors raise the possibility that shared protein deficits arise post-transcriptionally, via altered protein turnover or local synaptic regulation.

   Bulk RNA-seq, edgeR at FDR < 0.05: koFVB 23 genes up and 22 down, with RT-qPCR validation of Prnp, Plek, Igf1 (down) and Cplx1, Nrgn, Rph3a (up); ZH3 0 up and 2 down, with no PCA separation of KO from WT. The post-transcriptional mechanism is not tested. · Synaptic pathways are disrupted in PrPC…; Conserved glutamatergic and synaptic-marker…; Shared and model-specific consequences of PrPC…; Figure legends