Probing molecular mechanisms of neurodegenarative ageing in prion ablated mice

The projects at Roslin and Cardiff aimed to test the hypothesis that loss of PrP function causes degeneration of a discrete subset of neurons, detectable through molecular and physiological changes in the brains of mice lacking PrP or mice expressing mutated, functionally impaired, forms of PrP. We further hypothesised that this phenotype would be more prevalent during ageing.

We set up an ageing experiment involving 4 lines of mice (PrP-/-; PrP-P101L; a PrP-glycosylation deficient mouse; wildtype mice) and culled groups at time points from 300 days of age to 800 days. We studied transcriptome and proteome changes in two different groups of mice –those culled at 400 days old and those culled at 700 days old – by dissecting the brain longitudinally and performing microarray and proteomic analyses on the same animals. In 400 day old PrP knockout mice, we identified a range of differentially expressed genes/proteins as a result of PrP gene ablation. These molecules include several previously shown to be involved in neurodegenerative phenotypes, including CD44, various heat shock proteins, GFAP and a-synuclein. A network involving many of these proteins has been generated by data analysis through Ingenuity Pathway Analysis and is associated with neurological disease. Also evident in PrP-null mice are disturbances in proteins/genes associated with crucial metabolic and biochemical pathways, including oxidative phosphorylation, glycolysis/glucogenesis and ion homeostasis; certain cytoskeletal proteins are also differentially expressed. These data represent the first demonstration of molecular changes resulting from PrP knockout and help to explain some of the phenotypes associated with PrP-null mice.

We also performed analogous post genomic experiments on gene-targeted transgenic mice expressing PrP with a P101L mutation (associated with human prion disease) or lacking glycosylation. Although in many cases the differentially expressed molecules were different from PrP-/- mice, similar pathways were affected. These mice also show molecular evidence consistent with neurodegeneration and perturbations in mitochondrial / energy functions and the data represent intriguing evidence that loss of function of PrP may play a key role in prion diseases. We studied post genomic changes in 700 day old mice of each genotype, relative to wildtype controls, and found a substantially smaller set of differentially expressed molecules. Thus, we propose that molecular alterations in our PrP knockout or mutant mice are consistent with advanced aging, resulting in a phenotype that is neurodegenerative; by later life, differences between genotypes will therefore largely be masked.

Finally, to determine whether the molecular neurodegenerative phenotype observed was indicative of neuronal loss, we sectioned brains and performed histological examinations. In 700 day old mice, we found vacuolation consistent with aging, but this was not different between any of the mouse models and was not significantly different from wildtype controls. We also investigated whether loss of PrP function was associated with focal increases in GFAP staining, but found no differences between mouse models. In no sections did we observe either specific or non-specific loss of neurons that was different from wildtype mice.

In this project we have achieved all objectives; the nature of the aging experiment involved inevitably resulted in us producing data into the final stages and we are currently compiling all data for thorough bioinformatic analyses in advance of publications. Our conclusions are that molecular differences consistent with a neurodegenerative phenotype are evident in the three lines of PrP mice studied, but that there is no corresponding neuronal loss. Thus, during TSE disease, loss of PrPC function may prime neurons for neurodegeneration, but the accompanying toxicity of PrPSc may be required for true neurodegeneration.