The plasminogen activator (PA) system plays an important role in cell adhesion, migration and invasion, and may require the coordinated expression of various proteins. The human urokinase-type plasminogen activator (uPA) receptor (uPAR) is a central protein component of the PA system. By binding its ligand uPA, uPAR can direct proteolysis of the extracellular matrix. Also, it is now apparent that uPAR can initiate proteolytic independent signal transduction to influence angiogenesis, inflammation, wound repair and tumour progression. To determine whether any novel proteins interacted with uPAR, a yeast two-hybrid screening analysis was undertaken using alternate uPAR domain constructs as baits. These included full-length three domain uPAR (uPAR-DIDIIDIII), two domain uPAR (uPAR-DIIDIII), and each individual uPAR domain (uPAR-DI, uPAR-DII and uPAR-DIII). A number of proteins were identified as putative candidate interactors for the alternate constructs, with two of special interest for uPAR-DIDIIDIII. These were the heat shock protein Mrj, and the extracellular matrix protein fibulin-2. The protein Mrj was shown to bind uPAR both in vitro and in vivo using GST-pull down and co-immunoprecipitation assays respectively. The GST-pull down assay identified the interaction between Mrj and uPAR dependent on the C-terminal domain of Mrj and DI of uPAR. Using in vivo co-immunoprecipitation analysis, Mrj also bound to uPAR. Preliminary data suggest the association between uPAR and Mrj may play a role in the regulation of apoptosis. In regard to the uPAR interactor of fibulin-2, a calcium dependent binding interaction with uPAR was identified using the GST-pull down assay. However due to the large molecular weight and stringent conditions needed to solubilise fibulin-2, it was not possible to co-immunoprecipitate both uPAR and fibulin-2. Together, the identification of both Mrj and fibulin-2 amongst other candidate interactors of uPAR presented here provides further insight into the intricate relationship between uPAR and other proteins which may influence a range of biological functions.

GABAB receptors are G-protein coupled receptors for the major inhibitory neurotransmitter in the mammalian central nervous system, γ-aminobutyric acid (GABA). The receptor is linked to a variety of disorders including epilepsy, pain, spasticity, drug addiction and cognitive impairment and is, therefore of major importance for drug discovery. The most abundant receptor isoforms GABABR1a and R1b differ by the presence in R1a of a pair of Nterminal extracellular complement control protein modules (CCP1 and CCP2) which - in other proteins - are generally involved in mediating specific protein-protein recognition. The CCP1 module contains disulphides but is natively disordered. In the current work, the yeast two-hybrid system was used to confirm an interaction of CCP1 of GABABR1a with the extracellular protein fibulin-2. Further work with the yeast twohybrid system extablished the novel interaction of the abundant extracellular matrix protein laminin, with GABABR1a CCP1, via its laminin globular (LG) domains. The laminin interaction was further characterised by surface plasmon resonance, demonstrating that several different domains are involved in the binding to the GABAB receptor CCPs. The primary binding site is located on laminin α5 LG4-5, but the E10 domains of the β1 chain and LG1-3 on α1 may also be involved. The pharmacological properties of the GABABR1a and R1b isoforms were studied by transient expression in Xenopus laevis oocytes. It was demonstrated that the agonist baclofen, as well as the antagonist CGP55845, appear to be more potent at GABABR1b compared to GABABR1a. Intriguingly, when recorded in the precence of laminin, GABABR1b/R2 expressing oocytes exhibited an increased baclofen-evoked response while the response in GABABR1a/R2 was completely abolished. In conclusion, the work demonstrates that laminin is a binding partner for GABABR1a CCPs. Such an interaction between the metabotropic GABA receptor and the extracellular matrix may lie behind the recently reported roles of GABA in neuronal migration and the laying down of neuronal circuitry during the development of parts of the central nervous system.

Dans le système nerveux central, le récepteur inhibiteur GABAB est un archétype des RCPGs hétérodimériques. Il est composé en effet de deux sous-unités, la sous-unité GABAB1 (B1a ou B1b) qui lie l’agoniste et la sous-unité GABAB2 couplée aux protéines G. L’activation de ce récepteur a un effet antinociceptif bien établi concernant les douleurs aiguës mais son effet reste cependant très limité en cas de douleurs neuropathiques. Notre hypothèse est que son activation et sa signalisation peuvent être altérées par des protéines partenaires, aboutissant à des processus de désinhibition dans la moelle épinière en conditions de neuropathie. Nos résultats mettent en évidence le rôle de deux protéines partenaires qui sont surexprimées en conditions douloureuses et qui diminuent l’activation du récepteur GABAB via deux mécanismes différents. D’un part, la protéine cytosolique 14-3-3? induit la dissociation de l’hétérodimère B1b/B2. Cette action a lieu principalement dans les compartiments post-synaptiques. D’autre part, la fibuline-2, protéine de la matrice extracellulaire diminue l’activation de l’hétérodimère B1a/B2. Il s’agit cette fois préférentiellement d’une action dans les compartiments pré-synaptiques. Des stratégies anti-sens (siRNA anti-14-3-3? ou anti-fibuline-2) ou des peptides de compétition sélectifs de l’interaction B1b/14-3-3? permettent de potentialiser les effets antinociceptifs d’un agoniste du récepteur sur un modèle animal de neuropathie. L’ensemble de ces résultats suggèrent que l’état d’oligomérisation des RCPGs peut être modulé in vivo par des protéines partenaires endogènes impliquées dans le développement ou le maintien d’états pathologiques de sensibilisation à la douleur. / In the central nervous system, the inhibitory GABAB receptor is an obligate heterodimeric GPCR that requires the association between GABAB1 (B1a or B1b) and GABAB2 subunits. The heterodimeric GABAB receptor activation has a well-known antinociceptive action in acute pain but its effect appears limited in pathological states. Our hypothesis is that the GABAB activation and signaling could be altered by partner proteins, thus resulting in desinhibition processes in the spinal cord. In the present study, we investigated the role of two partner proteins overexpressed in neuropathic states which decrease GABAB activation through two different mechanisms. On the one hand, the cytosolic 14-3-3? protein induces the dissociation of the heterodimer B1b/B2. This effect occurs in post-synaptic compartments. On the other hand, fibulin-2, an extracellular matrix protein, which decreases the activation of the heterodimer B1a/B2 localized preferentially in presynaptic compartments. Anti-sens strategies (anti-14-3-3? or anti-fibulin-2 siRNA) or competing peptides specific of 14-3-3?/B1b interaction, potentiate the antinociceptive effects of GABAB agonist in an animal model of neuropathic pain. Taken together, our data suggest that GPCR oligomeric state can be modulated in vivo by endogenous partners proteins that are involved in the development and the maintenance of pain sensitization.