Des résultats d’expériences cliniques de reconstruction de l’os maxillaire faites à partir de la greffe d’une "pâte osseuse" gélifiée par l’ajout de facteur tissulaire ont été le primum movens de ce travail. Cette "pâte osseuse", faite d’os en poudre et de plasma enrichi en plaquette (PRP) à laquelle on ajoute du facteur tissulaire, est un modèle à la fois de la coagulation et de la régénération osseuse. Pour analyser des effets de la coagulation, nous avons utilisé un modèle connu : la culture primaire de cellules endothéliales (HUVEC). Les effets in vitro des facteurs de coagulation, dénommés protéases de la coagulation, pris séparément, ont été bien étudiés dans ces cellules, néanmoins aucune information sur l’effet combiné de ces protéases ou du plasma en coagulation n’était connue. Nous avons mesuré la "signalisation calcium" comme réponse cellulaire aux différents agents et ces mesures de la signalisation calcium ont été complétées par la mesure d’une autre réponse biologique, à savoir la sécrétion de cytokines pro-inflammatoires (IL-6 et IL-8). Pour l’étude de la régénération osseuse, la signalisation calcium a été mesurée sur une lignée d’ostéosarcomes humains (SaOS-2), stimulée par des protéases de la voie extrinsèque de la coagulation (facteur VIIa, facteur Xa et thrombine). Comme réponse biologique complémentaire, nous avons évalué l’effet des protéases d’intérêt sur l’apoptose induite par l’absence de sérum dans le milieu de culture. Les premiers travaux, réalisés sur les HUVEC, nous ont permis de montrer que le facteur Xa et la thrombine induisaient des signaux calcium différents sur ces cellules en mono couches, alors que le complexe facteur tissulaire – facteur VIIa ne provoquait aucune signalisation calcium. Nous avons également pu montrer une addition des signaux calcium induits par le facteur Xa et la thrombine. L’activation in situ du facteur Xa et de la thrombine à partir de leur zymogène a permis à la fois de confirmer les résultats précédents et de se rapprocher de l’in vivo. Finalement, au plus proche de l’in vivo, les expériences faites avec du plasma en coagulation ont également permis de détecter un signal calcium. La réponse biologique (sécrétion d’IL-6 et d’IL-8) en aval du signal calcium a confirmé les résultats calcium. En ce qui concerne la régénération osseuse étudiée à partir de SaOS-2, nous avons démontré l’expression du facteur tissulaire sur la lignée SaOS-2 et nous avons montré que le facteur VIIa, le facteur Xa et la thrombine induisaient tous des signaux calcium. Ces signaux présentaient des caractéristiques propres suivant la ou les protéase(s) utilisée(s) pour la stimulation. Les mesures ont également permis de caractériser, sur ces cellules, les récepteurs activés par les protéases d’intérêt, à savoir les "protease-activated receptors" 1 et 2 (PAR-1 et PAR-2). Comme réponse biologique, nous avons mesuré la diminution de l’apoptose induite par les protéases en absence de sérum dans le milieu de culture. Il a ainsi été montré que seule l’activation du récepteur PAR-1 permettait de diminuer l’apoptose. Finalement, nous avons caractérisé la voie suivie, qui passait par la phosphoinositide 3-kinase et la voie des MAPK Raf/MEK/ERK 1/2. En conclusion, cette thèse a permis de montrer, d’une part, que le facteur Xa et la thrombine provoquent des réponses calciques et proinflammatoires additifs dans les cellules endothéliales et, d’autre part, que le complexe facteur tissulaire – facteur VIIa, le facteur Xa et la thrombine induisent des signaux calcium caractéristiques dans les ostéosarcomes par l’activation des récepteurs PAR, l’activation de PAR-1 diminuant l’apoptose induite par l’absence de sérum dans le milieu de culture.

Thromboxane A2 receptor (TP) has been shown to play important roles in multiple aspects of cancer development including regulation of tumor growth, survival and metastasis. Molecular mechanisms of TP mediated cancer cell invasion remain to be identified. TP agonist, I-BOP, significantly elevated several matrix metalloproteinases (MMPs) including MMP-1, MMP-3, MMP-9 and MMP-10 in A549 human lung adenocarcinoma cells overexpressing TPα (A549-TPα) or TPβ (A549-TPβ). Signaling pathways of I-BOP-induced MMP-1 expression were examined in further detail as a model system for MMPs induction. Signaling molecules involved in I-BOP-induced MMP-1 expression were identified by using specific inhibitors including small interfering (si)-RNAs of signaling molecules and promoter reporter assay. The results indicate that I-BOP-induced MMP-1 expression is mediated by protein kinase C (PKC), extracellular signal-regulated kinase (ERK)-activator protein-1(AP-1) and ERK-CCAAT/enhancer-binding protein β (C/EBPβ) pathways. I-BOP-induced cellular invasiveness of A549-TPα cells was blocked by, GM6001, a general inhibitor of MMPs. Knockdown of MMP-1 and MMP-9 by their respective siRNA partially reduced I-BOP-stimulated A549-TPα cells invasion suggesting that other MMPs induced by I-BOP were also involved. Furthermore, secreted MMP-1 in conditioned media from I-BOP-treated A549-TPα cells (CM-I-BOP) autocrinely induced monocyte chemoattractant protein-1 (MCP-1) expression. The induction of MCP-1 by MMP-1 in A549 cells was via activation of protease-activated receptor 2 (PAR2) instead of commonly assumed PAR1. This conclusion was reached from the following findings: (1) expression of MCP-1 induced by trypsin, a PAR2 agonist, was inhibited by a PAR2 antagonist. (2) expression of MCP-1 induced by MMP-1 and by CM-I-BOP was blocked by a PAR2 antagonist but not by other PAR antagonists; (3) expression of MCP-1 induced by MMP-1 and by CM-I-BOP was attenuated significantly by pretreatment of cells with PAR2-siRNA. Finally, MCP-1 also can be induced by direct activation of TP in a SP1 involved mechanism. CM-I-BOP enhanced MCP-1-dependent migration of RAW 264.7 macrophages. Co-culture of A549 cells with RAW 264.7 macrophages induced expression of MMPs, VEGF and MCP-1 genes, and increased the invasive potential in A549 cells. My studies provide molecular mechanisms by which TP-mediated cancer cell invasion and suggest that TP is a potential anti-cancer drug target.

Le Syndrome de l'Intestin Irritable (SII) est caractérisé par des douleurs abdominales chroniques, des diarrhées, des constipations ou l'alternance des deux. Les neurones sensitifs des patients seraient surexcitables et ainsi à l'origine de l'hypersensibilité viscérale ressentie. Parmi les médiateurs impliqués dans le SII, les protéases ont un rôle prépondérant notamment car il a été observé une augmentation de leur activité dans des surnageants de biopsies coliques de patients atteints du SII. Ces protéases sont capables d'activer les " Protease-Activated Receptors " (PARs) dans les neurones sensitifs de souris. L'activation des neurones sensitifs murins peut être déclenchée selon un mécanisme dépendant de PAR2 et générant une hypersensibilité viscérale, mais inhibée par un mécanisme dépendant de PAR4. Dans certaines cellules (plaquettes), les PARs ont des rôles très différents entre la souris et l'homme. L'expression et la signalisation des PARs dans les neurones sensitifs humains étant jusqu'alors inconnues, cette thèse a eu pour objectif général d'étudier la signalisation des PARs et d'identifier les effets de médiateurs présents dans les tissus de patients atteints du SII, sur leur capacité à activer des neurones sensitifs humains. Le but ultime était d'identifier de potentielles cibles thérapeutiques pour le traitement de la douleur viscérale associée au SII chez l'homme. Le premier objectif a été de mettre au point la culture de neurones humains des ganglions spinaux dorsaux (GSD) thoraciques, contenant des projections coliques de neurones sensitifs. Une culture de neurones sensitifs d'une pureté de 90% a été obtenue, ces neurones étaient viables et fonctionnels, mobilisant du calcium intracellulaire en réponse à différents agonistes pro-nociceptifs. L'expression des PARs a été étudiée dans ces cultures et comparée à celle dans des tissus de GSD frais. PAR1, PAR2 et PAR4 de même que d'autres récepteurs canaux de la famille des " Transcient Receptors Potential " (TRPs) étaient exprimés de façon comparable dans les GSD fraichement isolés et dans les cultures de neurones, validant ainsi le phénotype physiologique des neurones en culture. L'effet des agonistes des PARs a ensuite été étudié sur la signalisation calcique dans ces cultures de neurones sensitifs humains. De tous les agonistes peptidiques des PARs utilisés, seul le peptide activateur de PAR1 a conduit à une augmentation du flux calcique dans les neurones sensitifs humains, tandis que le peptide agoniste de PAR4 l'a réduite. Comme le peptide activateur de PAR1, la thrombine, une protéase capable d'activer à la fois PAR1 et PAR4 a induit une augmentation du flux calcique dans les neurones sensitifs humains. La mobilisation calcique induite par la thrombine a été inhibée par un antagoniste de PAR1 et potentialisée par un antagoniste de PAR4. La thrombine exerce donc un double effet contradictoire sur les neurones sensitifs humains, induisant un signal calcique par l'activation de PAR1 et l'inhibant par l'activation de PAR4. Les surnageants de biopsies coliques de patients atteints du SII mais pas ceux de contrôles sains, ont provoqué un flux calcique dans les neurones sensitifs humains par un mécanisme dépendant de l'activation de PAR1. Nos résultats mettent en évidence l'importance de réduire la sensibilisation des neurones sensitifs humains dépendante de l'activation de PAR1 dans le SII. Notre technique de culture de GSD humains offre de nouvelles applications en neurosciences et nos résultats permettent d'envisager de nouvelles perspectives thérapeutiques quant à l'utilisation d'un antagoniste de PAR1 pour le traitement de la douleur viscérale associée au SII. / Irritable Bowel Syndrome (IBS) is characterized by chronic abdominal pain, diarrhea and/or constipation. Sensory neurons from IBS patients are considered to be over-activated, being thereby responsible of the experienced visceral pain. Among mediators involved in IBS, proteases seem to have a prominent role. First, proteolytic activity is increased in the supernatants of colonic biopsies from IBS patients. Second, these proteases activate " Proteases-Activated Receptors " (PARs) in rodent sensory neurons. Rodent primary afferent activation can be induced by a PAR2-dependent mechanism, which generates visceral hypersensitivity. This mechanism can be inhibited by a PAR4-dependent mechanism. In some cell types such as platelets, PARs may have very different functions in rodent versus human. The expression and the signalization of PARs in human sensory neurons have never been explored. The general aim of this thesis was to study the signalization of PARs and to identify the effect of mediators present in tissues from IBS patients, on their ability to activate human sensory neurons. The final objective was to identify new possible therapeutic targets to treat visceral pain associated with IBS. The first objective was to establish a protocol for culturing human sensory neurons isolated from thoracic dorsal root ganglia (DRG). Thoracic DRG contain the stroma of sensory neurons projecting from the colon. The culture of human sensory neurons was 90% pure with viable and functional neurons, which were able to induce calcium flux in response to different pro-nociceptive agonists. PAR expression was studied in cultures and compared to fresh DRG tissues. PAR1, PAR2 and PAR4 as well as others receptors from the " Transcient Receptors Potential " (TRPs) channel family were similarly expressed in neurons from fresh DRG tissues and in cultured sensory neurons, thereby validating the physiological phenotype of human DRG neurons in culture. The effect of PAR agonists was studied by imaging calcium mobilization in human sensory neuron cultures. From all PAR agonist peptides used, only PAR1 agonist peptide was able to increase calcium signaling in human sensory neurons, while PAR4 agonist peptide was able to decrease this calcium signaling. Similar to PAR1 agonist peptide, thrombin a protease capable of activating both PAR1 and PAR4, induced an increase of calcium flux in human sensory neurons. Thrombin-induced calcium mobilization was inhibited by a PAR1 antagonist and was potentiated by a PAR4 antagonist. Thus, thrombin has a contradictory double effect on human sensory neurons, inducing calcium signaling through PAR1 activation and inhibiting calcium signaling through PAR4 activation. In addition, supernatants of colonic biopsies from IBS patients but not from healthy controls, provoked an increased calcium signaling in human sensory neurons. This effect was dependent on the activation of PAR1. In the context of IBS, our results highlight the importance of targeting PAR1-induced sensitization in human sensory neurons. The DRG culture technique we have established offers important and new applications in the domain of neurosciences. All together, these data point to new therapeutic possibilities for the use of PAR1 antagonist in the treatment of visceral pain associated with IBS.

Protease-Activated Receptors (PARs) belong to an unusual family of G Protein Coupled Receptors (GPCRs). Each of the four known members is activated by its own N-terminus exposed by proteolytic cleavage and there is no other endogenous agonist known to date. PAR2 is the second member of the family and it has been implicated in wide range of pathophysiological conditions, particularly in various inflammatory diseases and cancers. In contrast, very little is known about the PAR2 receptor itself despite having been discovered more than 10 years ago. The purpose of this project was to improve our understanding of PAR2 regulation by discovering new agonists and antagonists and using them to probe the structural and functional properties of the receptor. Chapter 1 provides a brief literature overview of the initial discovery of PAR2, what is known about the mechanism of receptor activation, information on the structures and properties of current agonists and an antagonist for PAR2, and the putative physiological roles of human PAR2. As well, it summarizes the aims of this thesis. Chapter 2 investigates the regulation of gene expression by two different agonists of PAR2, a synthetic hexapeptide, 2f-LIGRLO-NH2, and the endogenous activator, trypsin, the idea being that genes up- or down- regulated by both agonists may more accurately profile PAR2-selective events. The effects of PAR2 activation on gene transcription in the human kidney HEK293 cell line were studied using a DNA microarray consisting of 19,000 human genes in an attempt to broadly cover the human genome and associated cell pathways with PAR2 activation. About 2,500 genes were regulated similarly by both agonists and, for genes expressed more than 5-fold, the mRNA results were further analyzed by quantitative RT-PCR techniques. PAR2 activation was shown to be associated with cellular metabolism, cell cycle, mitogen-activated protein kinase pathways, histone deacetylase and sirtuin enzymes, inflammatory cytokines and anti-complement function. Chapter 3 described a range of molecular events surrounding the activation of the receptor. PAR2 mRNA expression was quantitated by qRT-PCR and cross-checked with an intracellular Ca2+ assay. In this way whole cell PAR2 could be correlated with cell surface expression of PAR2. Three cell lines expressing high levels of PAR2 were chosen for subsequent experiments, these being colorectal carcinoma HT29, lung carcinoma A549 and human embryonic kidney HEK293 cells. Receptor activation, internalization, desensitization and resensitization assays were carried out on these cell lines to define some key functions relevant for investigating inhibitors in subsequent chapters. Chapter 4 reports a PAR2 mutagenesis study designed to identify the location of the binding site on PAR2 for a specific peptide agonist. A homology model of PAR2 based on bovine rhodopsin was used for docking of an agonist ligand, and the docking results were then investigated via two successive rounds of PAR2 mutagenesis in which the effect of each mutation (20 in all) was separately investigated by changes in agonist potency in the intracellular calcium release assay. Five PAR2 mutants showed more than a 5-fold reduction in agonist potency, while three others showed up to a 7-fold reduction. Mutations found to be important for agonist activity were mapped back to the model. Because there was extensive clustering of these key mutated amino acids, it is likely that this study has pinpointed the precise binding site of the agonist peptide in PAR2. Interestingly, this site is within the transmembrane region of the receptor. Chapter 5 reports the design, discovery and development of novel PAR2 agonists and antagonists and their regulatory effects in a diverse array of cell types. Structure-activity relationships were used to examine influences on the first, sixth and seventh positions of a PAR2 agonist peptide. At least five compounds were found herein to be equiopotent with the most potent PAR2 agonist reported. Knowledge obtained from this study was then used to create the first non-peptidic agonists for PAR2. The most potent nonpeptidic agonist (retaining one natural amino acid) was at least equipotent with the best peptide agonists. Conversion to nonpeptidic antagonists proved to be successful and this chapter reports the most potent known nonpeptide antagonist, which was selective for PAR2 and active at low micromolar concentrations. It inhibited intracellular Ca2+ release induced by different PAR2 agonists (trypsin, 2f-LIGRLO-NH2, nonpeptide agonists) in multiple cell lines (HT29, Panc-1, A549, MKN1, MKN45, MDA-MB-231, HUVEC) that have been physiologically associated with PAR2. It also inhibited release of inflammatory cytokines IL-8 and IL-6 and shows antiproliferative activity against primary human cells. The antagonist is competitive, reversible and surmountable (pA2 6.11). This thesis summarizes a large body of work that provides valuable molecular insights to PAR2 regulation, and lays the groundwork for rational design and development of novel nonpeptidic agonists and antagonists of PAR2 as potentially valuable pharmacological probes in vivo and as useful leads to development of therapeutics for inflammatory diseases and cancers.

Proteinase-activated receptor-2 (PAR-2) is a member of a family of G-protein-coupled, seven-transmembrane domain receptors that are activated by proteolytic cleavage. The receptor is expressed in a number of different tissues and potential physiological activators identified thus far include trypsin and mast cell tryptase. Acrosin, a trypsin-like serine proteinase found in spermatozoa of all mammals, was found to cleave a model peptide fluorescent quenched substrate representing the cleavage site of PAR-2. This substrate was cleaved with kinetics similar to those of the known PAR-2 activators, trypsin and mast cell tryptase. Acrosin was also shown to induce significant intracellular calcium responses in Chinese hamster ovary cells stably expressing intact human PAR-2, most probably due to activation of the receptor. Immunohistochemical studies using PAR-2 specific antibodies indicated that the receptor is expressed by mouse oocytes, which suggests that acrosin may play additional role(s) in the fertilization process via the activation of PAR-2 on oocytes.

No description available.

Platelets act as a fundamental component of the hemostatic process and their activation leads to the formation of a stable clot at the injured endothelium surface. Thrombin, as the important physiological agonist, activates platelets through protease-activated receptors (PARs). Protease-activated receptors are one of the major receptors in platelets and belong to the seven-transmembrane G-protein couple receptor family. Four protease-activated receptors are found, named as PAR1, PAR2, PAR3 and PAR4. Human platelets express PAR1 and PAR4 and murine platelets express PAR4 and PAR3 instead of PAR1. Thrombin activates PARs through a unique mechanism, involving the cleavage of N-terminus of PAR receptors and the newly exposed N-terminus acts as its own tethered ligand to bind and activate the receptor. In this study, we characterized a new PAR1 specific activating peptide (TFRRRLSRATR), generated from the c-terminus of human platelet P2Y1 receptor, and evaluated its biological function. This peptide activated platelets in a concentration-dependent manner, causing shape change, aggregation, secretion and calcium mobilization. Its activation is completely inhibited by using BMS200261, a PAR-1 specific antagonist. Its specificity to PAR1 receptor is further confirmed by using TFRRR-peptide-pretreated washed platelets and murine platelets. The shape change induced by 10 microM peptide was totally abolished by Y-27632, an inhibitor of p160ROCK which is the downstream signal of G12/13 pathways. The TFRRR-peptide, YFLLRNP, and the physiological agonist thrombin selectively activated G12/13 pathways at low concentrations and began to activate both Gq and G12/13 pathways with increased concentrations. Similar to SFLLRN, the TFRRR-peptide caused phosphorylation of Akt and Erk in a P2Y12 receptor-dependent manner, and p-38 MAP kinase activation in a P2Y12-independent manner. The effects of this peptide are elicited by the first six amino acids (TFRRRL) whereas the remaining peptide (LSRATR), TFERRN, or TFEERN had no effects on platelets. Beside thrombin, PARs also can be activated by other proteases. Previous studies in our lab show that plasmin, a major extracellular protease, activates both human and murine platelets through prototypical cleavage of PAR4 (Quinton et al., 2004). In this study, we continue our study and investigate the molecular basis for the differential activation of murine and human platelets by plasmin. Plasmin-induced full aggregation is achieved at lower concentrations (0.1 U/mL) in murine platelets as compared to human platelets (1 U/mL). In COS7 cells expressing the murine PAR4 (mPAR4) receptor, 1 U/mL plasmin caused a higher intracellular calcium mobilization than in cells expressing the human PAR4 (hPAR4) receptor. This difference was reversed when the tethered ligand sequences of mPAR4 and hPAR4 were interchanged through site-directed mutagenesis. This difference between human and murine PAR4 is not because of the cofactor effect of PAR3 in murine platelets by showing that in both transfected cell lines and platelet system, PAR3 inhibits plasmin-induced PAR4 stimulation. All of the data suggest that murine platelets are more sensitive to activation by plasmin than human platelets due to differences in the primary sequence of PAR4. In contrast to thrombin-dependent activation of platelets, wherein PAR3 acts as a co-receptor, mPAR3 inhibits plasmin-induced PAR4 activation. Abnormal platelet activation causes thrombus formation and induces pathological conditions including stroke and atherosclerosis. Antithrombotic therapy is a widely used therapeutic method for stroke. However, currently used agents based on the irreversible inhibition of the platelet cyclooxygenases 1 and 2 or inhibition of P2Y12 receptors can cause unexpected bleeding or resistant side effects. Antithrombotic therapy targeting thrombin signaling is one of the new treatments under investigation and PAR1 antagonists are now in clinical trials. In this study, we investigate the effect of one of thrombin receptors, protease-activated receptor 4 (PAR4) in mice transient middle cerebral artery occlusion/ reperfusion (tMCAO/R) model. Our data show that PAR4 -/- mice have more than 80% reduction in infarct volume and significant improved neurological and motor function after 1 h MCAO followed by 23 h reperfusion. Examination of cellular responses to tMCAO/R indicates that PAR4-/- mice have less cellular death. Platelet/endothelial and leukocyte/endothelial interactions have been shown to play a critical role in the inflammatory responses during cerebral ischemic/reperfusion injury. Comparing wild-type with PAR4-/- mice platelets/endothelial and leukocyte/endothelial interactions, deficiency of PAR4 causes a significant decrease in both platelet/endothelial and leukocyte/endothelial interactions. In addition, PAR4-/- mice attenuate blood-brain barrier (BBB) disruption during tMCAO/R. All the data suggest that deficiency of PAR4 will protect against brain ischemic injury though attenuation of cerebral inflammatory responses including inflammatory cells extravasation and BBB disruption. Protease-activated receptor 4 (PAR4) is the only thrombin receptor existing in both human and murine platelets. The data we get in this study also have a beneficial effect for human study and inhibition of PAR4 may provide a novel potential therapeutic strategy for ischemic injury. / Physiology

La barrière hémato-encéphalique (BHE) contrôle le passage des médicaments, en partie par la présence d’ATP Binding Cassette (ABC) transporteurs. Dans de nombreuses pathologies cérébrales, la BHE est altérée. Parmi elles, les hémorragies intracérébrales (HIC), qui sont un effet iatrogène des anticoagulants. Des analyses cliniques montrent que les patients sous Anticoagulants Oraux Directs (AODs) présentent moins d’HIC que les patients traités avec les anticoagulants de référence, les anti-vitamine K (AVK), sans que les mécanismes cellulaires soient élucidés. Une des différences entre les AODs et les AVK résident dans leur profil pharmacocinétique, effectivement, les AODs sont des substrats des ABC transporteurs contrairement aux AVKs. Au cours des HIC, la thrombine est activée et entraine une altération de la BHE par clivage et des récepteurs protease activated receptor (PAR). Les objectifs de ce travail de thèse ont été de mettre en place un modèle in vitro de BHE afin d’étudier les interactions des médicaments avec les ABC transporteurs. Ensuite, le modèle est utilisé pour étudier les interactions des AODs en condition pathologique. Le modèle développé est basé sur la lignée HBEC-5i, peu décrite dans la littérature. Les cellules ont été cultivées en monocouche sur insert avec milieu conditionné issu d’astrocytes humains. Le modèle permet l’étude de l’interaction de thérapeutiques avec des ABC transporteurs par des mesures d’efflux ratios. Le modèle a été validé par des études de transport de molécules pharmacologiques. Ensuite, nous avons comparé, sur notre modèle, les effets de l’exposition à la thrombine avec ou sans prétraitement d’anticoagulants (rivaroxaban, dabigatran, apixaban, warfarine et héparine). Les AODs limitent l’ouverture de la BHE induite par la thrombine contrairement aux autres anticoagulants. Nos résultats ont montré que l’altération de la BHE est médiée par le clivage du récepteur PAR-1 par la thrombine. Ce clivage n’est pas le même en fonction de la classe d’anticoagulants utilisée, les AODs minimisant ce clivage. L’ensemble de ce travail de thèse a permis de donner des premières explications cellulaires quant aux mécanismes d’ouverture de la BHE consécutifs aux HIC sous AODs. / The blood-brain barrier (BBB) controls the passage of drugs, in part through the expression of the ATP Binding Cassette (ABC) transporters. In many brain diseases, the BBB is altered. Among them, intracerebral haemorrhages (ICH), which are an iatrogenic effect of anticoagulants. Clinical analyses show that patients with Direct Oral Anticoagulants (DOACs) treatment have less HIC than patients treated with the reference anticoagulants, Vitamin K Antagonist (VKA), without understanding the cellular mechanisms. One of the differences between DOACs and VKA lies in their pharmacokinetic profile, indeed, DOACs are substrates of ABC transporters unlike VKA. During HIC, thrombin, is activated and causes alterations in the BBB by the cleavage of the protease activated receptor (PAR). The objectives of this thesis work were first to set up an in vitro model of the BBB in order to study the passage of drugs and their interactions with ABC transporters. In a second step, the model is used to study the interactions of DOACs in pathological conditions. The model developed is based on the HBEC-5i cell line seldom described in the literature. The cells were cultured in monolayer on insert with conditioned medium from human astrocytes. It allows the study of the interaction of therapeutics with ABC transporters by measuring efflux ratios. The model has been validated by transport studies of pharmacological molecules. In order to meet our second objective, we compared the effect of thrombin exposure with or without pretreatment with anticoagulants (rivaroxaban, dabigatran, apixaban, warfarin and heparin sodium) on our model. DOACs limit the BBB damage induced by the thrombin unlike other anticoagulants. Our results showed that alteration of the BBB is mediated by the cleavage of the PAR-1 receptor by thrombin. This cleavage is not the same depending on the class of anticoagulants used, DOACs minimizing this cleavage. All this thesis work made it possible to provide the first cellular explanations of the opening mechanisms of the BBB following HIC under DOACs.

Inflammation involves a complex interplay of immunological responses in reaction to an infection from invading pathogens and to tissue injury. However, if the inflammatory stimulus is not eliminated or if there is an aberrant response, chronic inflammation and disease can result. Most inflammatory diseases are targeted by drugs that provide only symptomatic relief by reducing the associated pain and swelling, without combating the underlying cause and progression of the disease. The purpose of this thesis was to investigate potential drug targets in different sub-cellular compartments to learn more about inflammatory pathways and the consequences of inhibiting them using different classes of small-molecule agents. Chapter 1 is a brief overview of inflammation, outlining differences between acute and chronic inflammation in relation to normal and aberrant immune responses. Major drug classes that are currently in use are briefly discussed, and some potential new inflammatory drug targets in the nucleus, cytoplasm, plasma membrane and serum that are to be investigated in the thesis are described. Chapter 2 addresses histone deacetylase (HDAC) enzymes that modulate gene transcription, and investigates the regulatory potential of HDACs and their inhibitors in macrophages. Mainly murine macrophages were investigated, but preliminary data on human macrophages is also presented. Four kinds of small-molecule inhibitors of HDAC enzymes are described in relation to their capacity to exhibit pro- or anti- inflammatory gene expression. Broad-spectrum inhibitors, those that do not discriminate between different isoforms of HDACs, were found to induce expression of a number of pro-inflammatory mediators. This was also the case for compounds that selectively inhibit class-I HDACs. On the other hand, compounds that are known to selectively inhibit class-II HDACs did not regulate pro-inflammatory genes and could be classified as anti-inflammatory based on their profiles of gene expression. An inhibitor of class-III HDACs, the sirtuins, was also briefly investigated for anti-inflammatory gene expression. This separation of the pro- and anti-inflammatory responses suggested that specific inhibitors of class II HDACs might be promising for treating inflammatory conditions. Chapters 3 and 4 discuss a group of extracellular serum proteins that are produced through the complement pathway. Complement proteins circulate in plasma while some are formed on membranes through serine protease action, and effect both innate and adaptive immune responses. Chapter 3 deals with the alternative pathway (AP) of complement activation, in particular with the protease Factor B which plays an important role as part of a protein complex (C3 convertase) in the initiation and propagation of complement. Factor B is an inactive zymogen at pH 7, but is catalytically active at alkaline pH. Substrate-based inhibitors were developed for Factor B using a chromogenic assay previously optimised in our lab. Over 60 peptide aldehydes were evaluated, with two potent inhibitors of Factor B studied further, and were found to inhibit C3 convertase by blocking cleavage of the native substrate C3 as measured by SDS-PAGE. Chapter 4 examined the properties of a homologue of Factor B found in the classical pathway (CP), the protease C2. C2 was found herein to be an inactive zymogen at pH 7. Under alkaline conditions C2 cleaved both C3 and short chromogenic peptide substrates. Substrates of 7- or 8-residues, corresponding to those in the native substrate C3, were optimally processed. C2 was inhibited by inhibitors identified in Chapter 3, which also inhibited formation of the different C3 convertase formed through the CP, as well as preventing formation of the terminal membrane attack complex (MAC) as measured by an immobilized ELISA and haemolysis assay. The selectivity of small-molecule peptidic inhibitors for different serine proteases was investigated. Chapter 5 introduces two other unrelated inflammatory proteins which have been incompletely studied herein. Protease-activated receptor 2 (PAR-2) is a cell surface protein that spans the membrane like other G-protein coupled receptors. PAR-2 has been implicated in inflammatory and proliferative diseased states, though its precise roles are still unclear. Here, PAR-2 mRNA expression was measured for different human cancer and immune cells. In macrophages, PAR-2 mRNA expression was inducible by HDACi, indicating that PAR-2 is negatively regulated by HDAC enzymes. The receptor was present and functional on the surface of HDACi-treated macrophages, permitting evaluation of PAR-2 agonists and antagonists. Caspase-1 is a cytosolic cysteine protease responsible for cleaving proIL-1β and proIL-18 into their active forms that are then secreted from cells as pro-inflammatory cytokines. Using monocytic cells, novel inhibitors of caspase-1 were developed and found to dose-dependently decrease levels of IL-1β released by the cell. Inhibition of caspase-1 may be useful for modulating circulating levels of the inflammatory cytokine IL-1β. The investigation of different prospective inflammatory targets in different locations of the cell has provided a lateral overview of inflammatory mechanisms and drug intervention strategies. Specific inhibitors have also been identified for these targets; HDACs in the nucleus, caspase-1 in the cytosol, PAR-2 and MAC on the cell surface, C2 and Factor B and C3 convertase in plasma. This unconventional study of the inflammatory network provides a different viewpoint of inflammatory proteins as drug targets.

O receptor tipo 2 ativado por protease (PAR2) é um receptor pro-inflamatório que pode ser ativado por tripsina, triptase, protease 3 produzida pelo neutrófilo e pela gingipaína (produzida por Porphyromonas gingivalis). Objetivo: o objetivo do presente estudo foi investigar a expressão do PAR2 na periodontite crônica em humanos, e avaliar se esta expressão está relacionada com a presença de atividade proteolítica no fluido crevicular. Metodologia: foram coletadas amostras de fluido crevicular gengival de indivíduos do grupo controle (sítios saudáveis com profundidade de sondagem &#8804; 3mm, ausência de sangramento à sondagem; n=40), grupo periodontite crônica com destruição de leve a moderada (n=40) e grupo periodontite crônica com destruição avançada (n=40). A expressão do PAR2 foi determinada por RT-PCR e a atividade proteolítica no fluido crevicular foi analisada utilizando-se o substrato BApNA. Resultados: observou-se um aumento significativo (p<0.001) da expressão do PAR2 no grupo periodontite em relação ao grupo controle, independentemente da severidade da doença. Além disso, no grupo periodontite houve um aumento significativo na atividade proteolítica (p<0.001) presente no fluido crevicular comparado com o grupo controle. Conclusões: concluiu-se que na periodontite crônica, há um aumento da expressão do PAR2 e um aumento da atividade proteolítica do fluido crevicular. Desta forma, sugere-se que o PAR2 pode estar envolvido com a inflamação periodontal em humanos. / Proteinase-activated receptor-2 (PAR2) is a pro-inflammatory receptor that can be activated by trypsin, tryptase, neutrophil-serine protease-3 (P3), and gingipain (produced by Porphyromonas gingivalis). Objectives: the objective of the present study was to investigate the expression of PAR2 in chronic periodontitis, and to evaluate whether its expression is related to the presence of proteolytic activity at the crevicular fluid. Methodology: gingival crevicular fluid (CF) samples were collected from subjects from control group (health sites with probing depth &#8804; 3mm, absence of bleeding on probing; n=40), chronic periodontitis group with slight to moderate destruction (n=40), and chronic periodontitis group with advanced destruction (n=40). PAR2 expression was determined by reverse transcriptase-PCR, and the proteolytic activity (PA) at the CF was quantified using the specific substrate BApNA. Results: PAR2 expression was significantly higher (p<0.001) in periodontitis compared to healthy individuals, independently of the diseases severity. In addition, in periodontitis group, there was a significantly increased proteolytic activity (p<0.001) compared with controls. Conclusions: we conclude that in chronic periodontitis there is an increased expression of PAR2 and an increased proteolytic activity at the crevicular fluid. Thus, it is suggested that PAR2 might be involved in human periodontal inflammation.