Pathways and targets in models of seeded aggregation

A commonality among many neurodegenerative disorders is the occurrence and progression of specific aggregated proteins in the brain. The most common neurodegenerative proteopathy is Alzheimer’s disease (AD), in which the aggregation and seeded propagation of amyloid-β peptide (Aβ) triggers a pathogenic process that includes neuronal Tau inclusions and neurodegeneration. It is not known when pathogenic Aβ seeds begin to form, propagate, and spread through the brain, nor has the biochemical nature of the initial Aβ seeds been defined. I shall summarize recent findings from the literature and in transgenic mouse models suggesting the presence of pathogenic Aβ seeds already during the lag-phase of protein aggregation in brain. At least one antibody used in clinical trials was identified to recognize such early seeds, which may differ from disease end-stage seeds. Thus, the preclinical phase of AD – currently defined as β-amyloid presence without clinical symptoms – may be a relatively late manifestation of a much earlier pathogenic seed formation and propagation that currently escapes detection in vivo.

Filamentous tau pathology is a hallmark feature of several neurodegenerative diseases, collectively known as tauopathies. Amongst others, P301L and N279K mutations in the gene encoding for tau (MAPT) have been linked to hereditary frontotemporal dementia (FTDP-17T). Recently, several animal and in vitro models have been generated to investigate disease mechanisms in FTDP-17T. However, these models do not fully recapitulate all the aspects of the disease that can be possibly identified in a human system. To investigate the alterations associated with MAPT mutations, we initially set up 2D cultures of induced-pluripotent stem cell (iPSC)-derived neurons from patients with the P301L and N279K mutations, and healthy controls (Iovino et al., 2015). The same IPSCs have now been used to create forebrain and midbrain organoids – brain regions affected in FTDP-17T, to determine whether the 3D models could give better information on the effect of the mutations in different brain regions besides providing a useful human system to investigate tau pathology and spreading. We have been able to generate cerebral organoids with regional identity, which contain both neuronal and glial cells. For both types of organoids, we performed a detailed histological and biochemical characterisation of the tau species and their modifications at different time points. In some organoids, we could observe alterations similar to those seen in tauopathy human brain. In conclusion, these organoids can be used as a tool to investigate mechanisms of tau toxicity and spreading in a human system and for the discovery of novel therapies for neurodegenerative diseases.

The hypothesis that aggregated tau can spread from one neuron to a second synaptically connected neuron is now a leading hypothesis for the mechanism of progression of tauopathies. Specific receptors have been identified that are responsible for tau uptake into neurons, however the cellular mechanisms that control the uptake of tau aggregates from the extracellular space and the pathways through which tau aggregates escape from the endosomal/lysosomal system to seed new tau aggregates in the cytoplasm of neurons remain largely unexplored.
To identify new genes and pathways involved in tau seeded aggregation, we performed three genetic screens. The first, a genome-wide CRISPR screen using HEK 293 cells that expressed a venus-tagged tau transgene seeded with heparin aggregated human tau employed a novel FACS method to sort cells based on tau puncta. The second, a screen in rat primary neurons that only expressed endogenous levels of tau and was seeded using soluble tau aggregates purified from Alzheimer’s disease brains, prioritized screening genes selected based on human genetics. The third, using primary neurons from tg4510 transgenic mice to amplify the aggregation signal observed in other systems, was deployed to validate hits in a different experimental system. Genes and pathways validated in these screens will be presented here.

Although alpha-synuclein is implicated in the pathogenesis of Parkinson’s disease and related disorders, it remains unclear whether specific conformations or levels of alpha-synuclein assemblies are toxic and how they cause progressive loss of human dopaminergic neurons. To address this issue, we used iPSC-derived dopaminergic neurons where endogenous alpha-synuclein was seeded with fibrils generated de novo or amplified from homogenates of brains affected with Parkinson’s disease (n=3) or multiple system atrophy (n=5). We showed that progressive neuronal loss in this model is dependent on alpha-synuclein levels and the conformation induced by specific strains. Transcriptomic analysis and isogenic correction of alpha-synuclein levels revealed the central role of alpha-synuclein in triggering neuronal death. We used proximity-dependent biotinylation in living cells and identified 56 differentially interacting proteins with endogenously assembled alpha-synuclein. We found that aggregates triggered with brain amplified fibrils evaded the Parkinson’s disease-associated deglycase DJ-1 and DJ-1 knockout enhanced aggregate-induced cell death in human dopaminergic neurons. Our results define parameters for iPSC-based modelling of alpha-synuclein pathology using disease-relevant fibrils and demonstrate how Parkinson’s disease-associated genes influence the phenotypic manifestation of strains in human neurons.

The pathogenesis of Parkinson's disease (PD) and Multiple system atrophy (MSA) involves the accumulation of aggregated forms of α-Synuclein in the brain. Progressive staging of α-Synuclein pathology in patient brains correlates with disease symptoms, and disease heterogeneity may be caused by distinct α-Synuclein strains. α-Synuclein pathology can spread in the brain of rodents following intracerebral injection of pathological α-Synuclein seeds. In the current studies, we investigated the ability of recombinant α-Synuclein assemblies with different conformations, i.e. fibrils and ribbons, as well as PD and MSA derived assemblies, i.e. Protein Misfolding Cyclic amplification (PMCA) assemblies and sarkosyl-insoluble fractions, to induce α-Synuclein pathology in primary neurons and mouse brains in vivo.
We showed that recombinant α-Synuclein assemblies induced seeding, with higher potency of fibrils compared to ribbons. PD and MSA PMCA assemblies were potent seeders, although distinct regional and morphological patterns of α-Synuclein pathology were observed in vivo. Sarkosyl-insoluble samples from MSA, but not from PD, were potent seeders in primary neurons.
Ubiquitin colocalization with α-Synuclein aggregates is a robust neuropathological hallmark in patient brains and a better understanding of the regulation of ubiquitinylation may provide important knowledge about the cellular response to α-Synuclein accumulation. We used the primary neuron seeding model to identify specific deubiquitinase enzymes which upon silencing reduced α-Synuclein seeding. These enzymes are under further investigation.
Taken together, our findings support that distinct α-Synuclein strains may define different synucleinopathies and that in vitro and in particular in vivo seeding models might be useful for validating new potential treatment principles.

A prevailing hypothesis in the field of neurodegenerative diseases, and Parkinson’s Disease (PD) in particular, is that pathogenic protein conformations may seed physiological endogenous proteins, leading to further protein aggregation and disease propagation. Consistent with this notion, the seeding ability of alpha-synuclein (AS) Pre-Formed Fibrils (PFFs) on endogenous AS has been well demonstrated in various cellular and animal systems, yet the mechanisms that lead to the generation and dissolution of such aggregates are unclear. Accordingly, we have applied human AS PPFs to a neuronally differentiated human neuroblastoma cell system with inducible tetracycline-dependent overexpression of human Wild Type (WT) AS. We observe time- and PFF dose-dependent seeding of endogenous AS in the Tet-off state, seeding that can be demonstrated with various AS antibodies. Antibodies directed against phosphorylated AS identified seeding only with much higher PFF concentrations. Experiments using pharmacological agents suggested that the lysosomal, and in particular the macroautophagic pathway, was responsible for the clearance of seeded WT AS, while the proteasomal system was involved specifically in the degradation of seeded phosphorylated AS. Clearance of seeded protein aggregates may also occur through extracellular release, although this could potentially lead to disease transmission. We examined accordingly if we could detect seeded AS load in released extracellular vesicles termed exosomes. Indeed, exosomes derived from PFF-treated Tet-off cells contained abundant seeded aggregated AS; however, levels of such species within exosomes did not change following lysosomal or proteasomal inhibition. We conclude that exosomal release represents another possible mechanism for the clearance of seeded aggregated AS.