Expression of the cellular prion protein (PrPC) from the PRNP gene is crucial for the development of a group of fatal neurodegenerative disorders called prion diseases. During prion infection a misfolded protein homologue of PrPC, PrPSc causes further misfolding on interaction with native PrPC molecules. PrPSc is highly resistant to proteinase K and aggregation of this protein is considered a hallmark of infection. Sheep are considered a model of natural infection and susceptibility to scrapie in sheep is defined by polymorphisms in the PRNP gene. It is still not fully understood how these polymorphisms regulate the conversion process or which other co-factors are involved. One such factor may be the truncation of PrPC via proteolytic processing in the form of two main cleavage events, known as α- and β-cleavage. In sheep α-cleavage cuts at amino acid 115, creating two truncated proteins C1 and N1 and represents the main cleavage event in healthy brain. β-Cleavage creates a longer C-terminal fragment, C2 and corresponding N-terminal fragment N2, cutting around amino acid 92 in sheep. Truncated forms of PrPC have been shown to represent around 50 % of total residual PrP in brain and may be an important determinant of disease through both decreasing the amount of full length PrPC available for conversion and through functions associated with the truncated fragments. The research presented has shown that increased production of an α-cleavage fragment C1 in brain is associated with TSE resistant genotype ARR/ARR, while the presence of C2 fragment is affiliated with scrapie susceptible PRNP genotypes in brain. There was no difference in the levels of full length PrPC in these genotypes suggesting that PrP expression does not directly correlate to susceptibility in this model. To assess if PrPC fragments could affect the conversion during disease in-vitro fibrillisation assays were performed using novel truncated recombinant proteins. These truncated proteins, although not thought to convert to PK resistant PrPSc during disease, can form amyloid fibrils. However, these fibrils appear to be less neurotoxic when compared to fibrils produced by full length PrPC. Only the truncated fragments derived from the ARR allele inhibit in-vitro fibrillisation of other allelic PrPC variants. Furthermore, treatment of infected cells in culture with recombinant C1ARR led to a decrease in the formation of disease associated PrPSc. In conclusion, genetic variations in levels of PrP truncated fragments may add to the complexity of genetic determinants of prion disease. In parallel with polymorphism-dependant conversion abilities, varying α-cleavage of ovine PrPC may help to explain genetic resistance in sheep. The inhibitory effects of C1, illustrated in-vitro may represent a therapeutic avenue in the treatment of prion disease.

The host encoded cellular prion protein (PrPC) is an N-linked glycoprotein tethered to the cell membrane by a glycophosphatidylinositol (GPI) anchor. Under certain conditions, PrPC can undergo conversion into a conformationally-altered isoform (PrPSc) widely believed to be the pathogenic agent of transmissible spongiform encephalopathies (TSEs). Thus, tissues expressing PrPC are potential sites for conversion of PrPSc during TSE pathogenesis. Although much is known about the role of PrPSc in prion diseases, the normal function of PrPC is poorly understood. Lines of mice and cattle in which PrPC has been ablated by gene knockout show no major phenotypical alterations other than resistance to TSE infection. However, recent reports using Prnp-null mouse models have suggested the participation of PrPC in neural stem/progenitor cell proliferation and differentiation. The first objective in our study was to map the expression of PrPC in twenty six somatic and reproductive tissues in ruminants. Our second objective was to characterize the ontogeny of PrPC expression during bovine embryonic and early fetal development. Finally, we used a mouse embryonic stem cell (mESC) model to study the potential role of PrPC during neurogenesis. In adult tissues, intense expression of PrPC was detected in the central nervous system (CNS), thymus and testes, whereas the liver, striated muscle and female reproductive tissues showed the lowest expression. We observed that PrPC was associated with tissues undergoing cellular differentiation including spermatogenesis, lymphocyte activation and hair follicle regeneration. Analyses in bovine embryos and fetuses indicated peaks in expression of PrPC at days 4 and 18 post-fertilization, stages associated with the maternal-zygote transition and the maternal recognition of pregnancy and initiation of placental attachment, respectively. Later in development, PrPC was expressed in the CNS where it was localized in mature neurons of the neuroepithelium and emerging neural trunks. Based on these observations, we hypothesized that PrPC was involved in neurogenesis. We tested this hypothesis in a murine embryonic stem cell model (mESC). mESC were induced to form embryoid bodies (EBs) by placing them in suspension culture under differentiating conditions and allowed to differentiate in vitro for 20 days. We detected increasing levels of PrPC starting on day 12 (8.21- fold higher vs. day 0; P < 0.05) and continuing until day 20 (20.77-fold higher vs. day 0; P < 0.05). PrPC expression was negatively correlated with pluripotency marker Oct-4 (r= -0.85) confirming that mESC had indeed differentiated. To provide a more robust system for assessing the role of PrPC in neural differentiation, mESC were cultured with or without retinoic acid (RA) to encourage differentiation into neural lineages. Induction of EBs with retinoic acid (RA) resulted in an earlier up-regulation of PrPC and nestin (day 12 vs. day 16; P < 0.05). In addition, immunofluorescence studies indicated co-expression of PrPC and nestin in the same cells. The results of these experiments suggested a temporal link between PrPC expression and expression of nestin, a marker of neural progenitor cells. We next tested whether PrPC was involved in RA-enhanced neural differentiation from mESC using a PrPC knockdown model. Plasmid vectors designed to express either a PrP-targeted shRNA or scrambled, control shRNA were transfected into mESC. Stable transfectants were selected under G418 and cloned. PrP-targeted and control shRNA clones, as well as wild-type mESC, were differentiated in presence of RA and sampled as above. PrPC expression was knocked down in PrP-targeted shRNA cultures between days 12 and 20 (62.2 % average reduction vs. scrambled shRNA controls). Nestin expression was reduced at days 16 and 20 in PrPC knockdown cells (61.3% and 70.7%, respectively vs. scrambled shRNA controls). These results provide evidence that PrPC plays a role in the neural differentiation at a point up-stream from the stages at which nestin is expressed. In conclusion, the widely distributed expression of PrPC in ruminant tissues suggests an important biological role for this protein. In the present work we have provided evidence for the participation of PrPC in the differentiation of mESC along the neurogenic pathway. / Ph. D.

Die genaue physiologische Funktion des zellulären Prion-Proteins (PrPC) ist noch immer nicht vollständig verstanden. Eine mögliche Funktion des PrPC auf das neuronale Überleben nach einem hypoxischen oder ischämischen Insult wird diskutiert. In einem Vorversuch zeigten sich nach zerebraler Ischämie deutlich größere Infarktvolumina in den Gehirnen von Prion-Knock-Out-Mäusen im Vergleich zu denen der Wild-Typ-Mäuse. Das Identifizieren der molekularen Mechanismen der PrPC-vermittelten Neuroprotektion ist daher von großem Interesse und machte die Etablierung eines Zell-Modells erforderlich. Neuere Studien konnten einen Einfluss des zellulären Prion-Proteins auf die Glykolyse nachweisen. Unter Sauerstoffmangelbedingungen kommt es zu einer vermehrten Bildung von Laktat durch das Enzym Laktat-Dehydrogenase (LDH). Neurone benötigen unter hypoxischen oder ischämischen Bedingungen dieses Laktat als Energiesubstrat. Je mehr Laktat den Neuronen zur Verfügung steht, umso höher ist das neuronale Überleben. In dieser Arbeit konnte die Beteiligung der Laktat-Dehydrogenase an der durch das zelluläre Prion-Protein vermittelten Neuroprotektion nach Hypoxie nachgewiesen werden. Das Ziel dieser Arbeit bestand darin, mögliche Unterschiede der LDH-Expression in WT-Zellen, Prnp0/0-Zellen und HEK-293-Zellen unter normalen und hypoxischen Bedingungen in vitro zu untersuchen. Die Expression der LDH war unter hypoxischen Bedingungen in den WT-Zellen im Vergleich zu den Prnp0/0-Zellen deutlich höher. Dies konnte auch in PrPC-überexprimierenden HEK-293-Zellen nach Hypoxie gezeigt werden. Ebenso konnte nachgewiesen werden, dass Hypoxie zu einem größeren Schaden des Tubulinzytoskelettes in Prnp0/0-Zellen führt als in WT-Zellen, was eine neuroprotektive Wirkung von PrPC vermuten lässt. Eine direkte oder indirekte Interaktion von LDH-A und PrPC konnte durch eine Co-Immunpräzipitation in HEK-293-Zellen nachgewiesen werden. Die genauen Mechanismen über die PrPC möglicherweise zu einer vermehrten Laktat-Produktion führt, sind noch nicht eindeutig identifiziert und müssen noch näher untersucht werden. Zusammengefasst kann gesagt werden, dass die erhobenen Daten die Vermutung verstärken, dass das Enzym LDH und sein Produkt Laktat in die durch das zelluläre Prion-Protein vermittelte Neuroprotektion nach Hypoxie involviert sind. Es ist das erste Mal, dass gezeigt wurde, durch welchen Mechanismus PrPC zur Neuroprotektion beiträgt.

Les mécanismes de type prion sont désormais reconnus comme sous-tendant la plupart des maladies neurodégénératives humaines, avec en premier lieu la maladie d’Alzheimer (MA) au niveau de ses 2 marqueurs spécifiques, l’amyloïde β (Aβ à l’origine de l’hypothèse étiopathogénique de la cascade amyloïde) et la protéine Tau phosphorylée. Par ailleurs la protéine du prion (PrPC) est décrite comme interagissant à de multiples niveaux avec le métabolisme de l’Aβ sans que les mécanismes physiopathologiques sous-jacents n’aient pu être expliqués. Pour sortir de l’impasse actuelle concernant le développement d’approches thérapeutiques efficaces pour la MA, l’industrie pharmaceutique a besoin de modèles expérimentaux innovants. En effet, à ce jour aucun modèle in vivo, en dépit des progrès réalisés avec les souris transgéniques, n’arrive à refléter la complexité cérébrale humaine ni à mimer une MA clinique. Les cultures in vitro en 2D sont quant à elles très éloignées des situations conduisant à l’accumulation d’agrégats protéiques pathologiques. Le but de notre thèse a été d’utiliser dans le domaine des neurosciences les nouvelles perspectives de recherche ouvertes par les technologies des cellules souches pluripotentes induites (cellules iPS) en développant un modèle de différentiation en 3D pour obtenir des organoïdes cérébraux humains (OC) (mini cerveaux). Leur capacité d’auto-organisation en 3D de tissu neuroectodermique nous a permis de recréer un système complexe mimant différentes structures cérébrales humaines dans lesquelles nous avons pu caractériser les marqueurs attendus. L’étude de l’expression des protéines d’intérêt APP et PrPC pendant la différentiation neurale a permis de caractériser la modulation des niveaux des deux protéines en fonction du temps de culture. Afin d’orienter le modèle vers des mécanismes d’accumulation protéique de type MA, nous avons testé différents inducteurs chimiques dont l’Aftin-5 qui est capable de moduler les voies post-traductionnelles de l’APP. Plusieurs stratégies de traitement ont été adoptées pour induire le clivage de l’APP et la génération d’Aβ. La production des fragments solubles Aβ38, Aβ40, Aβ42 a été mise en évidence par ELISA. Les niveaux générés sont reproductibles et l’augmentation du ratio Aβ42/Aβ40 est cohérente avec les données extrapolées des modèles murins et humains, ce qui a permis de valider notre modèle. Les niveaux d’expression génique et protéique de PrPC et de APP suite au traitement ont été analysés afin de mieux déterminer le rôle de l’interaction entre ces deux facteurs. L’objectif à long terme consiste à améliorer ce modèle, dont les limites actuelles sont notamment l’absence de vascularisation et le niveau de maturation du tissu neural. Le défi majeur dans le cadre de la culture des OC consiste donc à favoriser l’intégration du système vasculaire, et par ailleurs à accélérer le vieillissement in vitro pour l’étude de maladies neurodégénératives. La perspective de pouvoir automatiser le système de culture des OC permet d’envisager l’utilisation de ce modèle à plus grande échelle dans le cadre de test de cytotoxicité et/ou de criblage pharmacologique à haut débit pour identifier de nouvelles molécules thérapeutiques pour la MA. / Prion-like mechanisms are known to underlie most of human neurodegenerative diseases including Alzheimer’s disease (AD), which is characterized by two important pathological markers, β amyloid (or Aβ at the origin of the etiopathogenic amyloid cascade hypothesis) and phosphorylated tau protein. Furthermore, the prion protein (PrPC) interacts at multiple levels with the metabolism of Aβ, by mechanisms which are not well understood. To overcome the current limits in the development of efficient strategies to treat AD, the pharmaceutical industry requires innovative experimental models. However, even if a lot of progress has been achieved by using transgenic mouse models, to date no in vivo model can reflect the complexity of human brain or reproduce a clinical context. 2D in vitro cell culture models are unable to allow the aggregation and accumulation of pathological proteins as observed in vivo. The aim of this study consists in taking advantage of the research prospects offered by induced pluripotent stem cell (iPSCs) in the field of neurosciences. iPSCs can be used to generate 3D models of differentiation also called human cerebral organoids or mini-brains (MBs). Their ability to self-organise in 3D neuroectodermic tissue leds to a complex system that mimics different human cerebral structures in which we were able to characterize the expected markers. The study of the two proteins of interest (APP and PrPC) during neural differentiation has allowed us to follow the modulation of protein expression level occurring during the in vitro development of the human MBs. In order to use this model to reproduce the protein accumulation mechanisms seen in AD, we have tested chemical inductors such as Aftin-5 in order to modulate the APP post-transcriptional pathway towards a pathological outcome. Many strategies of treatment are adopted to lead APP cleavage and Aβ generation. The production of soluble fragments Aβ38, Aβ40, Aβ42 in the supernatant of organoids has been showed using ELISA technique. The levels generated are reproducible and the increase of Aβ42/Aβ40 ratio is consistent with extrapolated data from mouse and human models thus validating our model. Analysis at the gene and protein level has been assessed in order to understand the interaction between PrPC and APP after treatment. The long-term goal consists in improving this model which is notably hampered by the absence of vascularization and the low level of maturation of the neural tissue. The main challenge in MB culture thus consists in the integration of the vascular system, and also in increasing the speed of ageing process in vitro for the study of neurodegenerative diseases. In the long term, the prospect of automating the culture of MBs would allow the use of the system for cytotoxicity testing and/or high throughput screening for the discovery of new drugs for AD.