Cell type-specific translatome analysis of mouse models of three genetic neurodegenerative diseases

Sammanfattning: The burden neurodegenerative diseases place on patients, their loved ones, and the healthcare system is significant, and despite extensive research efforts, there is currently no cure. Since degenerative changes in the brain can begin years before symptoms appear, early intervention is critical. Additionally, neurodegenerative diseases target certain brain regions and neuron types early on. A more comprehensive understanding of the affected cells during the presymptomatic phase is therefore crucial for an effective and targeted intervention. Herein, we isolated, sequenced, and analyzed translatome samples from six neuronal cell types in knock-in mouse models of three monogenic neurodegenerative diseases at a presymptomatic stage: genetic Creutzfeldt-Jakob disease (gCJD), fatal familial insomnia (FFI), and Huntington’s disease (HD). To obtain the translatome samples, we used RiboTag to immunoprecipitate HA-tagged ribosomes with their translating mRNAs from targeted cell types. We analyzed six cell types across two brain regions: cerebral and cerebellar glutamatergic and GABAergic neurons, and cerebral parvalbumin (PV) and somatostatin (SST)-expressing neurons. In the first paper, we focused our analysis on the prion diseases, gCJD (E200K) and FFI (D178N). Here observed a similar response of SST+ neurons, a cell type not previously reported as affected, in both disease models. This was characterized by upregulation of ribosomeassociated genes, and downregulation of cytoskeleton and synapse-associated genes in FFI. Weighted gene co-expression network analysis of SST+ neurons pointed towards the downregulation of mTOR inhibition as a potential mechanism underlying the observed gene expression changes. In the second paper, we analyzed a 129S4-HdhQ200 knock-in mouse model of HD. Histological and behavioral assessment revealed pathological changes in the striatum and cerebellum at 9 months and a later, mild behavioral phenotype. Translatome analysis indicated a surprisingly strong response in reportedly resistant glutamatergic neurons of the cerebellum, marked by upregulation of cell cycle regulators Ccnd1 and chromobox protein genes. In the third paper, we aimed to compare disease-specific responses of PV+ neurons across the three disease models. This analysis revealed a milder response in HD compared to prion disease at comparable disease stages. Functional analysis further indicated PV+ neurons may respond differently in the investigated diseases, showing upregulation of immune response-associated pathways in gCJD, neurodegenerative-disease pathways in FFI, and autophagy in HD. Lastly, the generation of mouse models such as were used in papers I-III requires stable and predictable transgene expression without interfering with the expression of endogenous genes. In the fourth paper, we conducted a pilot study to compare three potential loci, Rpl6, Rpl7, and Eef1a1, as potential safe harbors for transgene integration. Preliminary results indicated that the Rpl6 locus may be best suited for our purposes. Furthermore, this work generated a novel dataset consisting of translatome profiles of six cell types in three neurodegenerative disease models. This provides gene expression data at a previously unavailable level of cellular resolution, especially in prion disease. We believe that this data will serve as a valuable resource for future research and help expand our understanding of the early molecular mechanisms in neurodegenerative disease beyond the scope of this thesis.