Neuronal membrane proteasomes (NMPs) are a functionally transmembrane subset of 20S proteasomes that degrade newly synthesized proteins. To date, the molecular composition of NMPs is undefined, and moreover, whether NMPs can influence any aspect of protein aggregation with relevance to neurodegenerative disorders remains unexplored. Using a Cre-dependent conditional knock-in mouse line to endogenously tag the proteasome, we find that NMPs co-purify with ApoE. We discover that NMP membrane localization is differentially modulated by ApoE isoforms (E4<E3<E2)in vitro,in vivo, and in human postmortem samples. This isoform-dependent change in NMP localization inversely correlates with the risk that ApoE isoforms pose for Alzheimer’s Disease. ApoE4-dependent reduction of NMP localization is strongest in brain regions selectively vulnerable to neurodegeneration. We synthesized selective NMP-specific inhibitors and discovered that NMP inhibition induces aggregation of endogenous and newly synthesized mouse and human Tau isoforms, without the need for seeding or pathogenic mutations. We posit that newly synthesized Tau is exceptionally susceptible to aggregation due to NMP dysfunction. Stereotactic injection of NMP inhibitorsin vivoinduces aggregation, phosphorylation, somatodendritic mislocalization and pathology of endogenous newly synthesized Tau. Finally, using ApoE-KI/hTau-KI crosses, we find that ApoE isoforms differentially shift the aggregation threshold for Tau. Overall, our data define NMPs as a pivotal proteostasis mechanism underlying the formation of endogenous Tau aggregates, which is directly regulated by the largest genetic risk factor for late-onset Alzheimer’s Disease.

Amyotrophic Lateral Sclerosis (ALS) is a fatal adult-onset neurodegenerative disease without cure or understanding of its molecular basis. It is characterized by progressive loss of spinal and upper motor neurons, but surprising resistance of cranial motor neurons until the late stages of the disease. There is no known single-gene cause of ALS, suggesting a failure of the entire cellular system to maintain proteostasis. To generate new hypotheses on the molecular differences between cranial and spinal motor neurons that underlie ALS-sensitivity, we exploited a unique platform that generates highly pure populations of induced cranial and spinal motor neurons (iCrMN and iSpMN). Exposing both iCrMNs and iSpMNs to proteotoxic stress, we monitored transcriptome and proteome changes over 36 hours and identified significant hundreds of statistically significant differences at either or both the RNA and protein level from a core set of 8,206 genes with complete data. While we observed a response to misfolding stress coordinated between transcript and protein changes, differences between the cell type were specific to RNA and protein expression. iCrMNs and iSpMNs differed in expression of many membrane proteins, including synaptic proteins, solute carriers, channels, and signaling molecules, as well as genes involved in ribosome biogenesis and subunits of the core proteasome. Differences exclusive to the transcriptome involved chaperones and the tricarboxylic acid cycle. Further, we found that the superior proteostasis capability of iCrMNs might, in part, be caused due to higher activity and abundance of the fully assembled 26S/30S proteasome in the nucleus, pointing to an increased role of ubiquitin-mediated, targeted protein degradation. The formation of the 26S/30S proteasome was highly dynamic under stress, likely mediated by a surplus of regulatory particles that attached and detached based on the action of proteasome regulators. Several lines of evidence supported this notion and described an intricate network of differentially expressed proteasome regulators and ubiquitination enzymes that defined the iCrMN- and iSpMN-specific ability to maintain proteostasis.