The yeast prion domain Sup35 NM models features of human neurodegenerative diseases in vivo

Abstract

In neurodegenerative diseases, such as prion diseases and Alzheimer’s disease, disease- associated proteins misfold and form amyloid deposits that progressively invade the CNS, leading to severe neurodegeneration. Accumulating evidence suggests that amyloid proteins propagate in a prion-like, self-perpetuating manner, but the mechanism of aggregate multiplication in mammals remains unclear. Amyloid deposits are associated with neurodegeneration and can induce a toxic gain-of-function or loss-of-function phenotype. Yet, the contribution of both effects to neurodegeneration is not fully understood. Surprisingly, the same pathogenic protein can aggregate into different structural variants that, similar to prion strains, may cause heterogenous clinical symptoms. However, how the amyloid structure can influence disease progression needs to be elucidated. Previously, our group established the NM-HA mouse model that expresses the hemagglutinin-tagged prion domain NM of the Saccharomyces cerevisiae prion Sup35 that behaves like a prion in mammalian cells. As NM does not possess a cellular function in mammals, it can be used to study the gain-of-function of protein aggregates in the absence of loss-of-function effects.<br /> In this project, we used the NM-HA mouse line to investigate the ability of fibril-induced NM-HA aggregates to propagate in the mammalian brain, and thus to model human prion-like proteins. Additionally, we tested if the gain-of-function of intracellular NM-HA aggregates can cause neurodegeneration and compared the disease pathogenesis induced by two different NM fibril conformers. Here we show that the intracranial injection of NM fibrils into NM-HA animals induces progressive NM-HA aggregation, demonstrating that yeast NM prions can replicate in mice. Interestingly, our data points to the involvement of the chaperone valosin- containing protein (VCP) as potential NM prion disaggregase in this process. NM-HA aggregates seeded by the NM fibril variants spread from the hippocampus to similar neuroanatomically connected regions, with striking similarity to pathologies observed in wild type mice challenged with disease-associated protein aggregates. Fibril-injected animals develop mild cognitive decline, likely caused by neuronal loss in hippocampal subregions with prominent NM-HA deposition. Remarkably, neurodegeneration is accompanied by local microgliosis and astrogliosis. Hence, a non-mammalian and non-disease-related protein is able to cause neurodegeneration upon aggregation in mice, likely via a neurotoxic gain-of-function effect. As fibril-injected NM-HA mice recapitulate key aspects of the pathogenesis of human neurodegenerative disorders, our data argue that mechanisms underlying intracellular amyloid fragmentation, dissemination, and toxicity might be shared between amyloidogenic proteins.