My project has been involved in analysing a class of interneuron that expresses the calcium‐binding protein parvalbumin (PV). In my thesis, I will describe the application of a method that involves the local injection of Creactivated recombinant adeno-associated viruses (AAVs) into a transgenic mouse line that expresses Cre recombinase in PV positive cells. This will drive the expression of a transgene specifically in PV positive cells, at a specific brain region. In the first part of my project, I used this method to introduce the molecular trans-synaptic tracer proteins wheat germ agglutinin and tetanus toxin heavy chain specifically to PV positive neurons to visualize their postand pre‐synaptic connections, respectively. What I found is that while our technique of combining Cre-activated AAVs in transgenic mice has allowed specific labelling of neurons in a brain region and cell type specific manner, we could not definitively identify trans-synaptically traced neurons. In the second part of my project I have used these novel AAV‐based techniques in mice to introduce tetanus toxin light chain (TeLC) to PV neurons in the dentate gyrus. This has been previously used to functionally remove PV neurons from the CA1 of the hippocampus. This protein inhibits neurotransmitter release by cleaving the vesicle docking protein, VAMP2. The DG has been implicated in the separation of sensory inputs (pattern separation) which increases the resolution of the encoded memory and thereby assists in the accurate recall. The lateral inhibition of excitatory activity in the DG is believed to aid accurate encoding. Using our AAV method, I found that PV positive interneurons are required for spatial working and reference memory. Using a new behavioural assay that I developed, I could also show that these neurons are needed to enhance the resolution of spatial information. However, I also discovered that long term expression of TeLC could result in neuronal cell death. I have therefore demonstrated that local injection of Cre recombinase activated AAVs allows for a quick, versatile method of genetic manipulation, provided long term expression (greater than 2 months) is not required.

Ο νικοτινικός υποδοχέας της ακετυλοχολίνης είναι ο σημαντικότερος στόχος στην αυτοάνοση νόσο βαριά μυασθένεια. Στο μεγαλύτερο ποσοστό των ασθενών (85%) ανιχνεύονται αυτοαντισώματα έναντι του μυϊκού υποδοχέα της ακετυλοχολίνης, τα οποία μπορούν να οδηγήσουν σε διαταραχή της μεταγωγής του κινητικού ερεθίσματος. Σε αυτή την μελέτη στόχος είναι η κατασκευή και έκφραση των εξωκυτταρικών περιοχών των υπομονάδων α και β του ανθρώπινου μυϊκού υποδοχέα της ακετυλοχολίνης σε κύτταρα εντόμων μετά από επιμόλυνση με βακιλοϊό. Στα ίδια πλαίσια πραγματοποιήθηκε η σύνδεση των υπομονάδων α και β με πεπτιδικό συνδέτη για την δημιουργία του συγκαταμερούς β-α. Τα ανασυνδυασμένα αυτά πεπτίδια απομονώθηκαν και χαρακτηρίστηκαν σε διαλυτή μορφή. Σε επόμενο στάδιο έγινε η ακινητοποίηση τους σε υπόστρωμα σεφαρόζης και ο έλεγχος της ικανότητας πρόσδεσης αυτοαντισωμάτων από ορό μυασθενών με απώτερο στόχο τηv χρήση τους ως πιθανοί ανοσοπροσροφητές σε μια καινούρια θεραπευτική προσέγγιση για την μυασθένεια. / Myasthenia gravis (MG) is an autoimmune disease caused in the majority of patients (85%) by autoantibodies against the human muscle acetylcholine receptor (AChR),a post-synaptic ligand-gated ion-channel located at the neuromuscular junction.Autoantibodies against the AChR cause loss of the available and functional AChRs leading to muscle weakness and fatigability.An attractive therapeutic approach is the extracorporeal specific removal of the pathogenic autoantibodies using AChR-based immunoadsorbents. In this study, the N-terminal extracellular domains (ECD) of AChR the subunits α1 and β1 were cloned into baculovirus expression vectors and heterologously expressed using insect SF9 cells.Additionally, the two subunits were linked by the use of a flexible peptide linker to form the β1-α1 concatamer.The recombinant proteins were expressed as soluble polypeptides, purified and characterized.Furthermore, they were immobilised on sepharose beads in order to test them as immunoadsorbents using sera from MG patients. They were all found to bind anti-AChR autoantibodies, albeit to varying degrees. They, thus, pose as potential candidates for the therapeutic antigen-specific clearance of MG sera.

Durant le développement, l’innervation d’une zone précise du cerveau par certaines branches axonales est un mécanisme encore mal compris. Afin d’aborder cette question, je me suis intéressée aux axones rétiniens qui innervent deux cibles principales du système visuel : le corps genouillé latéral dorsal (CGLd) et le colliculus supérieur. J’ai étudié le rôle de la protéine de guidage Sémaphorine 6D et de son récepteur Plexine-A1 dans l’innervation spécifique du CGLd par les axones rétiniens. J’ai ainsi découvert que chez les souris Sema6D-/- et Plexine-A1-/-, le tractus optique (formé par les axones rétiniens) entre dans le CGLd au lieu de le contourner et certains axones rétiniens innervent des régions ectopiques de l’autre côté du tractus optique. De plus, l’analyse des souris simple ou double hétérozygotes indique que ces deux protéines interagissent avec un mécanisme dose-dépendant. Grâce à des expériences de perte et de gain de fonction par électroporation rétinienne in utero, j’ai pu montrer la nécessité de Sema6D et de Plexine-A1 dans la rétine pour l’innervation des axones rétiniens et ce via des effets non cellulaire autonomes. Ces résultats révèlent un mécanisme dose-dépendant dans lequel Sema6D et PlexineA1 interagissent et assurent une communication axone-axone permettant l’innervation précise du CGLd par une sous-population d’axones rétiniens. / During development, axons branch at precise points to innervate a specific brain target, yet the mechanisms at hand are still unclear. To address this question, I used retinal axons forming the optic tract that innervate two principal targets of the visual system: the dorsal lateral geniculate nucleus (dLGN) and the superior colliculus. I investigated the role of the guidance receptor Plexin-A1and its ligand Semaphorin-6D (Sema6D) in this targeting process. Here I highlight a new type of phenotype in Plexin-A1-/- or Sema6D-/- mice. In these mice, the optic tract enters in the dLGN instead of circumscribing it and some retinal axons innervate ectopic regions at the other side of the optic tract. Furthermore, the analysis of simple or double heterozygotes mice reveals that Plexin-A1 and Sema6D interact together with a dose-dependent effect. Using loss and gain of function experiments (via retinal in utero electroporation), I showed that both are necessary in the retina for proper retinal innervation through non-cell autonomous effects. All these results reveal for the first time a dose-dependent mechanism, in which Sema6D and Plexin-A1 interact together. They monitor axon-axon communication to allow the correct innervation of the dLGN by a subpopulation of retinal axons.