Plants have developed specific responses to external stimuli such as drought, cold, high salinity in soil, and precipitation in addition to internal developmental stimuli. These stimuli trigger signal transduction pathways in plants, leading to cellular adaptation. A signal transduction pathway is a network of entities that interact with one another in response to given stimulus. Such participating entities control and affect gene expression in response to stimulus . For computational purposes, a signal transduction pathway is represented as a network where nodes are biological molecules. The interaction of two nodes is a directed edge. A plethora of research has been conducted to understand signal transduction pathways. However, there are a limited number of approaches to explore and integrate signal transduction pathways. Therefore, we need a platform to integrate together and to expand the information of each signal transduction pathway. One of the major computational challenges in inferring signal transduction pathways is that the addition of new nodes and edges can affect the information flow between existing ones in an unknown manner. Here, I develop the Beacon inference engine to address these computational challenges. This software engine employs a network inference approach to predict new edges. First, it uses mutual information and context likelihood relatedness to predict edges from gene expression time-series data. Subsequently, it incorporates prior knowledge to limit false-positive predictions. Finally, a naive Bayes classifier is used to predict new edges. The Beacon inference engine predicts new edges with a recall rate 77.6% and precision 81.4%. 24% of the total predicted edges are new i.e., they are not present in the prior knowledge. / Master of Science

Plants activate an array of defense mechanisms upon pathogen attack. Systemic acquired resistance (SAR) is an induced disease resistance phenomenon deployed after infection by a necrogenic pathogen and is dependent on endogenous accumulation of salicylic acid. The objectives of my research were to characterize SAR in the crop plant, <i>Brassica napus</i> (canola), and study the effects of overexpressing genes involved in SAR on disease resistance. Biological induction of SAR using necrogenic Pseudomonas syringae and chemical induction using benzo (1,2,3) thiadiazole-7-carbothionic acid reduced growth of the bacterial pathogen P. syringae and the fungal pathogen Leptosphaeria maculans. This growth reduction was associated with an increase in transcript levels of pathogenesis-related (PR) genes, one of the characteristic features of SAR. Transgenic plants expressing a bacterial salicylate hydroxylase gene (NahG), were more susceptible to the above pathogens and were delayed in accumulating PR gene transcripts, indicating a need for SA accumulation for SAR in B. napus. Expression of two SAR genes from Arabidopsis, DEFECTIVE IN INDUCED RESISTANCE 1 (DIR1) and NON EXPRESSOR OF PATHOGENESIS-RELATED 1 (NPR1), in <i>B. napus</i> enhanced resistance against virulent P. syringae without SAR pre-treatments. Putative orthologs of DIR1 and NPR1 (BnDIR1 and BnNPR1) were isolated from B. napus based on EST sequences. BnDIR1 and BnNPR1 display 71% and 66% amino acid sequence similarities, respectively, to the corresponding Arabidopsis proteins. Expression of BnNPR1 in Arabidopsis npr1 mutant backgrounds indicated that it was able to functionally complement these mutations. Expression of BnDIR1 enhanced disease resistance in both Arabidopsis wild-type and dir1-1 mutant backgrounds. Expression of DIR1, NPR1, BnDIR1 and BnNPR1, separately, in <i>B. napus</i> plants enhanced resistance against P. syringae. SAR pre-treatments further enhanced resistance of transgenic <i>B. napus</i> plants expressing DIR1 and BnDIR1 to <i>P. syringae</i>, indicating an additive effect. Expression of DIR1 in B. napus did not provide resistance against <i>L. maculans</i>. These results provide the first in-depth molecular characterization of SAR in B. napus, and in particular, provide new insight into DIR1 function not previously reported in Arabidopsis.

Plants activate an array of defense mechanisms upon pathogen attack. Systemic acquired resistance (SAR) is an induced disease resistance phenomenon deployed after infection by a necrogenic pathogen and is dependent on endogenous accumulation of salicylic acid. The objectives of my research were to characterize SAR in the crop plant, <i>Brassica napus</i> (canola), and study the effects of overexpressing genes involved in SAR on disease resistance. Biological induction of SAR using necrogenic Pseudomonas syringae and chemical induction using benzo (1,2,3) thiadiazole-7-carbothionic acid reduced growth of the bacterial pathogen P. syringae and the fungal pathogen Leptosphaeria maculans. This growth reduction was associated with an increase in transcript levels of pathogenesis-related (PR) genes, one of the characteristic features of SAR. Transgenic plants expressing a bacterial salicylate hydroxylase gene (NahG), were more susceptible to the above pathogens and were delayed in accumulating PR gene transcripts, indicating a need for SA accumulation for SAR in B. napus. Expression of two SAR genes from Arabidopsis, DEFECTIVE IN INDUCED RESISTANCE 1 (DIR1) and NON EXPRESSOR OF PATHOGENESIS-RELATED 1 (NPR1), in <i>B. napus</i> enhanced resistance against virulent P. syringae without SAR pre-treatments. Putative orthologs of DIR1 and NPR1 (BnDIR1 and BnNPR1) were isolated from B. napus based on EST sequences. BnDIR1 and BnNPR1 display 71% and 66% amino acid sequence similarities, respectively, to the corresponding Arabidopsis proteins. Expression of BnNPR1 in Arabidopsis npr1 mutant backgrounds indicated that it was able to functionally complement these mutations. Expression of BnDIR1 enhanced disease resistance in both Arabidopsis wild-type and dir1-1 mutant backgrounds. Expression of DIR1, NPR1, BnDIR1 and BnNPR1, separately, in <i>B. napus</i> plants enhanced resistance against P. syringae. SAR pre-treatments further enhanced resistance of transgenic <i>B. napus</i> plants expressing DIR1 and BnDIR1 to <i>P. syringae</i>, indicating an additive effect. Expression of DIR1 in B. napus did not provide resistance against <i>L. maculans</i>. These results provide the first in-depth molecular characterization of SAR in B. napus, and in particular, provide new insight into DIR1 function not previously reported in Arabidopsis.

Gene regulatory networks (GRNs) provide a natural representation of relationships between regulators and target genes. Though inferring GRN is a challenging task, many methods, including unsupervised and supervised approaches, have been developed in the literature. However, most of these methods target non-context-specific GRNs. Because the regulatory relationships consistently reprogram under different tissues or biological processes, non-context-specific GRNs may not fit some specific conditions. In addition, a detailed investigation of the prediction results has remained elusive. In this study, I propose to use a machine learning approach to predict GRNs that occur in developmental stage-specific networks and to show how it improves our understanding of the GRN in seed development. I developed a Beacon GRN inference tool to predict a GRN in seed development in Arabidopsis based on a support vector machine (SVM) local model. Using the time series gene expression levels in seed development and prior known regulatory relationships, I evaluated and predicted the GRN at this specific biological process. The prediction results show that one gene may be controlled by multiple regulators. The targets that are strongly positively correlated with their regulators are mostly expressed at the beginning of seed development. The direct targets were detected when I found a match between the promoter regions of the targets and the regulator's binding sequence. Our prediction provides a novel testable hypotheses of a GRN in seed development in Arabidopsis, and the Beacon GRN inference tool provides a valuable model system for context-specific GRN inference. / Master of Science

Ca2+/calmodulin-dependent protein kinase type-II (CaMK-II) is a Ser/Thr protein kinase regulated by Ca2+ and Calmodulin. It is a highly conserved and broadly expressed enzyme and has a unique structure and dynamic regulation. It has the ability to remain active in the absence of Ca 2+ as a result of Ca2+ dependent autophosphorylation. CaMK-II phospliorylates proteins involved in neurotransmitter secretion, long term potentiation, cytoskeletal dynamics, gene transcription, and cell motility. To support existing and identify new intracellular roles of CaMK-II, potential binding partners were identified. This was accomplished by transfecting and purifying "FLAG-tagged" CaMK-II's (α, βE, δC, and δE). CaMK-II associated proteins were then identified using tandem mass spectrometry. Known binding partners were identified using this approach, including CaMK-II and calmodulin, verifying the approach's validity. Additionally several unexpected but interesting proteins were identified, including the gelsolin related actin binding protein, Flightless-I. Fli-I is an actin binding and capping protein that also functions as a transcriptional coactivator. The CaMK-II-Fli-I interaction was confirmed with endogenous (un-tagged) proteins. The association and localization of Fli-I are dependent on CaMK-II's activity state, although Fli-I is not a substrate of CaMK-II. When CaMK-II is inhibited, Fli-I translocates to the nucleus. Conversely when CaMK-II is artificially activated using a Ca2+ ionophore, Fli-I returns to the cytosol. The discovery of this reversible interaction epresents a potentially new CaMK-II regulated pathway and likely serves as a link between Ca2+ based signal transduction pathways and regulation of the actin component of the cytoskeleton and transcription.

No description available.

<p>Phosphatidylinositol (PI) is a phospholipid that constitutes only a minor component of eukaryotic membranes. However, they are critical in many fundamental cellular processes, such as signal transduction pathways, vesicular trafficking and actin cytoskeletal dynamics. PI is highly enriched in specific acyl chains at both the <em>sn-1</em> and <em>sn-2</em> positions, the major species being 1-stearoyl-2-arachidonoyl. Enzymes required for PI synthesis are believed to play a major role in this enrichment through the selective catalysis of specific substrates. We have studied several aspects of two enzymes involved in PI synthesis, Diacylglycerol kinase ε (DGKε) and CDP-Diacylglycerol synthases (CDS). We have studied the role of the ATP-binding motif of DGKε and showed that this enzyme is not only required for enzymatic activity, but substrate specificity and sub-cellular localization. We have also looked at the region adjacent to the catalytic site, containing a cholesterol recognition motif, and determined that this also affects the enzymes activity and substrate specificity. Finally, we have characterized the enzymatic properties of two CDS isoforms <em>in vitro</em> and demonstrated that these isoforms exhibit different substrate specificities. Taken together, our results serve to further our understanding of both DGKε and CDS1/2 and their roles in PI synthesis and enrichment with specific acyl chains.</p> / Master of Science (MSc)

Hintergrund und Zielsetzung: DMD ist die häufigste Form der Muskeldystrophie im Kindesalter und bis heute unheilbar. Sie wird durch das Fehlen des Proteins Dystrophin verursacht, welches verschiedene Signaltransduktionswege beeinflusst. Das Anliegen der Arbeit ist die Untersuchung und Modulation von Signaltransduktionswegen, die als alternative Therapiestrategie den Verlust von Dystrophin kompensieren könnten. Experimentelle Strategie: Für die Charakterisierung von Dystrophin nachgeschalteten Prozessen wurden mRNA-Expressionsanalysen in Muskelgeweben von DMD-Patienten und einem DMD-Brüderpaar mit einem infrafamiliär unterschiedlichen Verlauf der DMD durchgeführt. Aus diesen Expressionsdaten wurde erstmalig ein Petri-Netz entwickelt, welches Dystrophin mit in diesem Zusammenhang bisher unbekannten Signaltransduktionswegen verknüpft. Das Petri-Netz wurde auf Netzwerkintegrität und –verhalten mittels Invarianten- (INA) und theoretischen Knockout- (Mauritius Maps) Analysen untersucht. Durch beide Methoden läßt sich der maßgebliche Teilsignalweg bestimmen. In diesem Signalweg wurden die Proteinaktivität und die Genexpression durch siRNA, Vektor-DNA und chemische Substanzen in humanen SkMCs moduliert. Anschließend wurden die Proliferation und die Vitalität der Zellen sowie auch die Expression auf mRNA- und Protein-Niveau untersucht. Ergebnisse: RAP2B und CSNK1A1 waren in dem DMD-Brüderpaar differentiell exprimiert und konnten erstmalig in einem neuen, komplexen Signalweg in Zusammenhang mit Dystrophin nachgeschalteten Prozessen dargestellt werden. Mittelpunkt dieses Signalweges ist die De- und Aktivierung des Transkriptionsfaktors NFATc. Seine Zielgene umfassen neben anderen den negativen Proliferationsfaktor p21, das Dystrophin homologe UTRN und den Differenzierungsfaktor MYF5. Folglich würde ein Anstieg von UTRN eine unerwünschte Reduktion der Proliferationsrate von Myoblasten implizieren. Letzteres konnte bereits nachgewiesen werden und stellte das Motiv für weitere Studien dar. Jedoch zeigten siRNA- und Vektor-DNA-Experimente, daß NFATc nicht der ausschlaggebende Faktor für diese Zielgene ist. Die Substanzen Deflazacort (DFZ) und Cyclosporin A (CsA) wurden dagegen beschrieben, die Aktivierung von NFATc zu beeinflussen. Die Ergebnisse zeigten, daß beide Substanzen die Proliferation von Myoblasten erhöhen können. Die gleichzeitige Applikation von DFZ und CsA führte zu einem Anstieg der UTRN-Expression. Schlußfolgerung: Die Modulation der Proliferation und UTRN-Expression ist unabhängig von einander möglich. Entsprechend der Grundidee der Arbeit zeichnet sich eine neue Therapiestrategie ab, welche Dystrophin nachgeschaltete Prozesse einbezieht. / Background and aim: DMD is the most common muscular dystrophy in childhood and incurable to date. It is caused by the absence of dystrophin, what influences several signal transduction pathways. The thesis is interested in the investigation and modulation of signal transduction pathways that may compensate the lack of dystrophin as an alternative therapy strategy. Experimental strategy: To study Dystrophin downstream pathways the mRNA expression of DMD patients and two DMD siblings with an intra-familially different course of DMD were analysed in muscle tissue. On the basis of these expression data a Petri net was first developed implicating signal transduction pathways and Dystrophin downstream cascades. Invariant (INA) and theoretical knockout (Mauritius Maps) analyses were applied for studying network integrity and behaviour. Both methods provide information about the most relevant part of the network. In this part modulation of protein activity and of gene expression using siRNA, vector-DNA, and chemical substances were performed on human SkMCs. Subsequently, the cells were studied by proliferation and vitality tests as well as expression analyses at mRNA and protein level. Results: RAP2B and CSNK1A1 were differently expressed in two DMD siblings, and first are part of a signal transduction pathway implicating Dystrophin downstream processes. The central point of this pathway is the de- and activation of the transcription factor NFATc. Its target genes are, among others, the negative proliferation factor p21, the Dystrophin homologue UTRN, and the differentiation factor MYF5. Consequently, an increase in UTRN implicates an undesirably reduced myoblast proliferation rate. Latter was found in DMD patients and was target for further studies. But, siRNA and vector DNA experiments showed that NFATc is not the decisive factor for the target genes. Deflazacort and cyclosporin A are known to influence the activation of NFATc. The results first showed that both substances do induce myoblast proliferation. The use of deflazacort in combination with cyclosporin A resulted in an increase of UTRN expression. Conclusion: The modulation of proliferation and UTRN-expression independently of each other is possible. According to the basic idea of this study, a new therapeutic strategy becomes apparent, which considers Dystrophin downstream processes.

The area of systems biology evolved in an attempt to introduce mathematical systems theory principles in biology. Although we believe that all biological processes are essentially chemical reactions, describing those using precise mathematical rules is not easy, primarily due to the complexity and enormity of biological systems. Here we introduce a formal approach for modeling biological dynamical relationships and diseases such as cancer. The immediate motivation behind this research is the urgency to find a practicable cure of cancer, the emperor of all maladies. Unlike other deadly endemic diseases such as plague, dengue and AIDS, cancer is characteristically heterogenic and hence requires a closer look into the genesis of the disease. The actual cause of cancer lies within our physiology. The process of cell division holds the clue to unravel the mysteries surrounding this disease. In normal scenario, all control mechanisms work in tandem and cell divides only when the division is required, for instance, to heal a wound platelet derived growth factor triggers cell division. The control mechanism is tightly regulated by several biochemical interactions commonly known as signal transduction pathways. However, from mathematical point of view, these pathways are marginal in nature and unable to cope with the multi-variability of a heterogenic disease like cancer. The present research is possibly one first attempt towards unraveling the mysteries surrounding the dynamics of a proliferating cell. A novel yet simple methodology is developed to bring all the marginal knowledge of the signaling pathways together to form the simplest mathematical abstract known as the Boolean Network. The malfunctioning in the cell by genetic mutations is formally modeled as stuck-at faults in the underlying Network. Finally a mathematical methodology is discovered to optimally find out the possible best combination drug therapy which can drive the cell from an undesirable condition of proliferation to a desirable condition of quiescence or apoptosis. Although, the complete biological validation was beyond the scope of the current research, the process of in-vitro validation has been already initiated by our collaborators. Once validated, this research will lead to a bright future in the field on personalized cancer therapy.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal. / Protein-protein interactions are an integral component of signal transduction pathways. The interactions are mediated by modular domains which are present within the structure of the signalling molecule. These domains include PDZ, SH2, SH3, WW, PTB and LIM domains. hPTP1E is a protein-tyrosine phosphatase which contains within its primary structure a region with homology to Band 4.1 proteins, five PDZ domains and a catalytic domain. While the function of this PTPase remains unknown, the structure of hPTP1E suggest it may recruit several proteins into a multiprotein complex. In order to understand the role of hPTP1E, the protein interactions within its signalling cascade were examined. hPTP1E interacts with ZRP-1 and GEF-5.1 via its second PDZ domain and tuberin via its fourth PDZ domain in the yeast two-hybrid system. In order to characterize these proteins and their interactions, antibodies were generated against ZRP-1 and GEF-5.1. The antibodies which detected the antigen expressed in bacteria were purified by affinity chromatography. The antibodies raised against ZRP-1 and hPTP1E detected proteins of appropriate molecular weights in total cell extracts. HPTP1E is ubiquitously expressed whereas ZRP-1 is restricted to HeLa and MCF-7 cells among the cells tested. Unfortunately, antibodies against GEF-5.1 did not detect a protein of the predicted molecular weight in any of the cell extracts. Immunoprecipitation of hPTP1E fi-om cells overexpressing regions of ZRP-1 and tuberin with a hemaglutinin tag demonstrated the presence of an interaction between the phosphatase and tuberin in vivo. However, ZRP-1 and hPTP1E did not interact under these experimental conditions. Confirmation of the yeast two-hybrid results provides further support for a possible role of hPTP1E in the regulation of endocytosis. Additional molecules involved in the signalling pathways involving hPTP1E were identified by interaction trap. In one study, the proline rich amino terminus of ZRP-1 interacted with several clones encoding a segment of hCDC47 and NtZRP-p33, a clone containing an SH3 domain. The significance of these findings is unknown. HCDC47 is a minichromosome maintenance protein wich regulate DNA replication. Further, a clone called KIAA0769 containing the sequence of NtZRP-p33 depicts the typical structure for a scaffolding protein. Another yeast-two hybrid cDNA library screening using the CDC25 homology domain of GEF-5.1 did not detect an interaction with any GTPase but with 14-3-3E. 14-3-3 proteins are regulatory molecules which interact with various types of proteins by means of a phosphorylated serine residue. Mutational analysis demonstrated that the interaction is dependent on the second serine residue within the consensus sequence RSLSQG found in GEF-5.1. The primary structure of the open reading frame of GEF-5.1 was analyzed using profilescan. The software predicted the presence of several domains including a cNMP binding domain, a LTE domain, a PDZ domain, a rasassociated domain and a CDC25 homology domain. A family of guanidine nucleotide exchange factors may exist as clones KIAA0313 and T14G10 have the same structure. These results indicate a role for GEF-5.1 in Ras signalling pathways. Further, its activity may be regulated by the binding of cNMP molecules and 14-3-3E. The identification of ZRP-1 and GEF-5.1 interacting proteins as well as the analysis of the primary structure of GEF-5.1 have provided additional information about the function of hPTP1E. This cytoplasmic phosphatase may be involved in the regulation of processes such as transcription, DNA replication and. Further, an interaction between tuberin and hPTP1E suggests a role for this PTPase in the regulation of endocytosis. / La phosphorylation des protéines est une modification post-traductionelle fréquemment employée pour moduler la transmission des signaux intracellulaires. Il est nécessaire qu'un équilibre du niveau de phosphorylation soit maintenu pour le fonctionnement normal de la cellule sinon des maladies comme le cancer peuvent apparaître. Les enzymes responsables de la phosphorylation des protéines sont les protéines kinases tandis que les protéines phosphatases enlèvent les groupements phosphate. Les résidus phosphorylés dans les protéines sont certains résidus sérines, thréonines et/ou tyrosines. Les différentes enzymes sont classées en deux familles selon leur spécificité. Les protéine-tyrosine phosphatases (PTPase) sont elles-même regroupées dans deux familles selon leur localisation intracellulaire: les PTPases de type récepteur et les phosphatases cytoplasmiques. La structure des phosphatases de type récepteur inclus un domaine extracellulaire, un domaine transmembranaire et un (ou deux) domaine(s) catalytique(s). Les PTPases cytoplasmiques contiennent un domaine catalytique unique et généralement un/ ou des domaine(s) responsable(s) de leur localisation intracellulaire ou impliqué(s) dans des interactions protéine-protéine. Dans notre laboratoire, une phosphatase cytoplasmique dénommée liPTP1E par nous (et PTPL1, PTPBAS, FAP par d'autres) a été isolée. En plus de son domaine catalytique, cette protéine-tyrosine phosphatase contient 1 domaine de type "Band 4.1" qui est impliqué dans la localisation de la protéine à la membrane cellulaire via une interaction avec le cytoskelette, et 5 domaines PDZ. Ces domaines PDZ sont en général impliqués dans les interactions protéineprotéine. Plusieurs études récentes ont tenté de définir la fonction de hPTP1E. Sato et ses collègues ont isolé hPTP1E lors d'un criblage d'une librairie d'ADNc en utilisant le système des deux-hybrides dans la levure avec la partie cytoplasmique du récepteur Fas, comme appât. Ils ont aussi démontré que hPTP1E peut inhiber l'effet apoptotique de Fas. L'apoptose des cellules cibles qui est induit par les lymphocytes T cytotoxiques utiliserait le système Fas. De plus, Fas pourrait être associé à des maladies auto-immunes. En plus, hPTP1E pourrait jouer un rôle dans l'apparition de cellules resistantes aux effets de Fas tel que retrouvées dans les sarcomes de Kaposi chez les sidéens. Malgré des données convaincantes, il reste quand même des doutes quant à l'importance de hPTP1E dans ces maladies. Ainsi une étude publiée n'a pu démontrer une interaction entre les homologues de Fas et hPTP1E chez la souris. Depuis d'autres groupes étudiant les interactions de hPTP1E ont découvert plusieurs protéines qui interagissent avec celle-ci. La première, PARG, est membre de la famille des Rho-GAP, des protéines impliquées dans l'activation des GTPases de type Rho. L'interaction aurait lieu avec le 4ième domaine PDZ de hPTP1E. De plus, le domaine LEVI de RIL interagirait avec hPTP1E via ses 2ième et 4ième domaines PDZ. La fonction biologique de ces interactions n'a toutefois pas été déterminée à ce jour. Pour caractériser la fonction biologique de hPTP1E, nous avons utilisé le système des deux-hybrides de la levure pour identifier des protéines qui interagiraient avec les domaines PDZ de hPTP1E. J'ai ainsi identifié deux protéines nommés ZRP-1 et GEF-5.1, qui se lient à hPTP1E. ZRP-1 possède une structure semblable à celle de zyxin.Ces deux dernières protéines contiennent une région amino-terminale riche en résidus proline et 3 domaines de type LEM à l'extrémité carboxyl terminale. GEF-5.1, d'autre part démontre une homologie marquée aux GEFs de la famille CDC25 impliquées dans l'activation des GTPases de la famille Ras. Des anticorps ont été générés contre ZRP-1, GEF-5.1 et hPTP1E afin de fournir les outils nécessaires pour mieux caractériser ces différentes protéines. Ainsi, j'ai exprimé et purifié le troisième domaine LIM de ZRP-1 ainsi que le domaine PDZ de GEF-5.1, sous forme de protéines de fusion avec la glutathioneS-transferase (GST). Ces protéines ont servis d'antigène pour générer des anticorps chez le lapin. Des anticorps dirigés contre le deuxième domaine PDZ de hPTP1E étaient déja disponibles dans le laboratoire. Ces anticorps ont été purifiés sur une colonne d'affinité GST. Les anticorps anti-ZRP-1 et anti-hPTP1E détectent tous les deux des protéines du poids moléculaire attendu. HPTP1E est exprimé d'une facon ubiquitaire tandis que l'expression de ZRP-1 est plus restrainte parmi les cellules testées. Toutefois, les immunoglobulines dirigées contre GEF-5.1 ne détectent aucune protéine du poids moléculaire attendu dans un extrait cellulaire brut. Parallèlement, d'autres membres du laboratoire ont démontré une interaction entre la tuberine, le produit du gène TSC2, un oncogène impliqué dans la sclérose tubéreuse, et le quatrième domaine PDZ de hPTP1E. Afin de caractériser ces interactions in vivo, des immunoprécipitations de hPTP1E à partir de cellules dans lesquelles une région de ZRP-1 et/ou de la tuberine étaient surexprimé ont été conduites. Sous les conditions expérimentales utilisées, ZRP-1 n'a pas co-immunoprécipité avec hPTP1E. Cependant une interaction avec la tuberine a été détectée utilisant cette stratégie suggérant que HPTP1E pourrait jouer un rôle dans la modulation de l'endocytose. La structure de ZRP-1 inclus un domaine riche en proline qui n'est pas nécessaire pour son interaction avec hPTP1E mais qui pourrait interagir avec d'autres protéines en particulier avec des protéines contenant un/ ou des domaine(s) SH3. La moitié amino-terminale de ce domaine a été utilisé pour cribler une librairie d'ADNc par le système des deux-hybrides. Un clone appelé NtZRP-p33 contenant un domaine SH3 a été identifié. La conséquence biologique de cette interaction reste toutefois a être déterminée. Cependant, NtZRP-p33 possède une structure suggérant son implication dans la signalisation intracellulaire. Un deuxième criblage de la librairie d'ADNc a été initié pour caractériser les protéines impliquées dans le mécanisme de signalisation de hPTP1E. En utilisant le domaine de GEF-5.1 homologue à CDC25, des clones correspondants à la protéine 14-3-3 ont été isolés. Les protéines 14-3-3 forment une famille de protéines qui régularisent la fonction de plusieurs protéines. Leurs interactions se font via un residu sérine qui est phosphorylé. Des mutations du domaine catalytique ont démontré que l'interaction entre 14-3-3s et GEF-5.1 est dépendante du deuxième sérine de la séquence RSLSQG qui se retrouve immédiatement du coté carboxyl terminale du domaine GEF de la protéine GEF-5.1. Ces résultats suggèrent que l'activité de GEF-5.1 pourrait être modulée par la 14-3-38. En conclusion, les résultats expérimentaux présentés dans ce mémoire indique un rôle potentiel de hPTP1E dans plusieurs fonctions cellulaires. En s'associant à la tuberine, hPTP1E pourrait régulariser l'endocytose. Aussi, cette PTPase pourrait être impliquer dans le cycle cellulaire. Ras étant un activateur de la mitose, HPTP1E pourrait moduler l'activité de Ras par voie de GEF-5.1. Ainsi, hPTP1E pourrait agir comme proto-oncogène ou un gène suppresseur des tumeurs. Zyxin est une protéine qui se retrouve près des sites membranaires en association avec le cytoskelette. Puisque la structure de ZRP-1 et zyxin est semblable, ce dernier sert de modèle pour la fonction de ZRP-1. En collaboration avec hPTP1E, ces deux protéines pourrait régulariser la structure du cytoskelette.