Rho GTPases are critical regulators of actin cytoskeletal dynamics. The three most well characterized Rho GTPases, Rac1, RhoA and Cdc42 share a common inhibitor, RhoGDI. It is only recently becoming clear how upstream signals cause the selective release of individual Rho GTPases from RhoGDI. For example, our laboratory showed that diacylglycerol kinase zeta (DGKz), which converts diacylglycerol (DAG) to phosphatidic acid (PA), activates PAK1-mediated RhoGDI phosphorylation on Ser-101/174, causing selective Rac1 release and activation. Phosphorylation of RhoGDI on Ser-34 by PKCa has recently been demonstrated to selectively release RhoA, promoting RhoA activation. Here, I show DGKz is required for optimal RhoA activation and RhoGDI Ser-34 phosphorylation. Both were substantially reduced in DGKz-null fibroblasts and occurred independently of DGKz activity, but required a function DGKz PDZ-binding motif. In contrast, Rac1 activation required DGKz-derived PA, but not PDZ-interactions, indicating DGKz regulates these Rho GTPases by two distinct regulatory complexes. Interestingly, RhoA bound directly to the DGKz C1A domain, the same region known to bind Rac1. By direct interactions with RhoA and PKCa, DGKz was required for the efficient co-precipitation of these proteins, suggesting it is important to assemble a signalling complex that functions as a RhoA-specific RhoGDI dissociation complex. Consequently, cells lacking DGKz exhibited decreased RhoA signalling downstream and disrupted stress fibers. Moreover, DGKz loss resulted in decreased stress fiber formation following the expression of a constitutively active RhoA mutant, suggesting it is also important for RhoA function following activation. This is consistent with the ability of DGKz to bind both active and inactive RhoA conformations. Collectively, these findings suggest DGKz is central to two distinct Rho GTPase activation complexes, each having different requirements for DGKz activity and PDZ interactions, and might regulate the balance of Rac1 and RhoA activity during dynamic changes to the actin cytoskeleton.

Rho GTPases are critical regulators of actin cytoskeletal dynamics. The three most well characterized Rho GTPases, Rac1, RhoA and Cdc42 share a common inhibitor, RhoGDI. It is only recently becoming clear how upstream signals cause the selective release of individual Rho GTPases from RhoGDI. For example, our laboratory showed that diacylglycerol kinase zeta (DGKz), which converts diacylglycerol (DAG) to phosphatidic acid (PA), activates PAK1-mediated RhoGDI phosphorylation on Ser-101/174, causing selective Rac1 release and activation. Phosphorylation of RhoGDI on Ser-34 by PKCa has recently been demonstrated to selectively release RhoA, promoting RhoA activation. Here, I show DGKz is required for optimal RhoA activation and RhoGDI Ser-34 phosphorylation. Both were substantially reduced in DGKz-null fibroblasts and occurred independently of DGKz activity, but required a function DGKz PDZ-binding motif. In contrast, Rac1 activation required DGKz-derived PA, but not PDZ-interactions, indicating DGKz regulates these Rho GTPases by two distinct regulatory complexes. Interestingly, RhoA bound directly to the DGKz C1A domain, the same region known to bind Rac1. By direct interactions with RhoA and PKCa, DGKz was required for the efficient co-precipitation of these proteins, suggesting it is important to assemble a signalling complex that functions as a RhoA-specific RhoGDI dissociation complex. Consequently, cells lacking DGKz exhibited decreased RhoA signalling downstream and disrupted stress fibers. Moreover, DGKz loss resulted in decreased stress fiber formation following the expression of a constitutively active RhoA mutant, suggesting it is also important for RhoA function following activation. This is consistent with the ability of DGKz to bind both active and inactive RhoA conformations. Collectively, these findings suggest DGKz is central to two distinct Rho GTPase activation complexes, each having different requirements for DGKz activity and PDZ interactions, and might regulate the balance of Rac1 and RhoA activity during dynamic changes to the actin cytoskeleton.

Arterial endothelial cells (ECs) are subjected to pulsatile strain due to pressure changes in the cardiac cycle and this may play a significant role in vascular function in health and disease. Further, ECs differentially respond to different patterns of strain. There is much evidence that cyclic uniaxial strain results in a perpendicular orientation of ECs and their stress fibers, while no such alignment occurs in response to cyclic equaibiaxial stretch. It is unclear how cells and their stress fibers determine their specific response to particular spatiotemporal changes in the matrix, however. Given that ECs located at regions in the arterial tree prone to atherogenesis are non-aglined, while ECs in relatively healthy regions are oriented perpendicular to the principal direction of cyclic stretch, it is important to understand the mechanisms which regulate stretch-induced stress fiber alignment. The focus of this thesis was to develop realistic models to describe the dynamic changes in the organization of stress fibers in response to diverse spatiotemporal patterns of stretch. The model is based on the premise that stress fibers are pre-stressed at a ?homeostatic? level so that stress fibers are extended beyond their unloaded lengths, and that perturbation in stress fiber length from the homeostatic level destabilizes the stress fibers. A deterministic model described experimentally measured time courses of stress fiber reorientation perpendicular to the direction of cyclic uniaxial stretch, as well as the lack of alignment in response to equibiaxial stretch. In the case of cyclic simple elongation with transverse matrix contraction, stress fibers oriented in the direction of least perturbation in stretch. Model analysis indicated the need for a time-dependent stress fiber mechanical property, however. Thus, a stochastic model was developed that incorporated the concept that stress fibers tend to self-adjust to an equilibrium level of extension when they are perturbed from their unload lengths with the turnover of stress fibers. The stochastic model successfully described experimentally measured time courses of stress fiber reorganization over a range of frequencies. At a frequency of 1 Hz, stress fibers predominantly oriented perpendicular to stretch, while at 0.1 Hz the extent of stress fiber alignment was markedly reduced and at 0.01 Hz there was no alignment at all. Both the deterministic and stochastic models accurately described the relationship between stretch magnitude and the extent of stress fiber alignment in endothelial cells subjected to cyclic uniaxial stretch. Parameter sensitivity analyses for each model were used to demonstrate the effects of each parameter on the characteristics of the system response. In summary, the mathematical models were capable of describing stress fiber reorganization in response to diverse temporal and spatial patterns of stretch. These models provide a theoretical framework to elucidate the mechanisms by which adherent cells sense the characteristics of matrix deformation and describe a mechanism by which the cells can then adapt to such deformations to maintain mechanical homeostasis.

Rho GTPases are critical regulators of actin cytoskeletal dynamics. The three most well characterized Rho GTPases, Rac1, RhoA and Cdc42 share a common inhibitor, RhoGDI. It is only recently becoming clear how upstream signals cause the selective release of individual Rho GTPases from RhoGDI. For example, our laboratory showed that diacylglycerol kinase zeta (DGKz), which converts diacylglycerol (DAG) to phosphatidic acid (PA), activates PAK1-mediated RhoGDI phosphorylation on Ser-101/174, causing selective Rac1 release and activation. Phosphorylation of RhoGDI on Ser-34 by PKCa has recently been demonstrated to selectively release RhoA, promoting RhoA activation. Here, I show DGKz is required for optimal RhoA activation and RhoGDI Ser-34 phosphorylation. Both were substantially reduced in DGKz-null fibroblasts and occurred independently of DGKz activity, but required a function DGKz PDZ-binding motif. In contrast, Rac1 activation required DGKz-derived PA, but not PDZ-interactions, indicating DGKz regulates these Rho GTPases by two distinct regulatory complexes. Interestingly, RhoA bound directly to the DGKz C1A domain, the same region known to bind Rac1. By direct interactions with RhoA and PKCa, DGKz was required for the efficient co-precipitation of these proteins, suggesting it is important to assemble a signalling complex that functions as a RhoA-specific RhoGDI dissociation complex. Consequently, cells lacking DGKz exhibited decreased RhoA signalling downstream and disrupted stress fibers. Moreover, DGKz loss resulted in decreased stress fiber formation following the expression of a constitutively active RhoA mutant, suggesting it is also important for RhoA function following activation. This is consistent with the ability of DGKz to bind both active and inactive RhoA conformations. Collectively, these findings suggest DGKz is central to two distinct Rho GTPase activation complexes, each having different requirements for DGKz activity and PDZ interactions, and might regulate the balance of Rac1 and RhoA activity during dynamic changes to the actin cytoskeleton.

Rho GTPases are critical regulators of actin cytoskeletal dynamics. The three most well characterized Rho GTPases, Rac1, RhoA and Cdc42 share a common inhibitor, RhoGDI. It is only recently becoming clear how upstream signals cause the selective release of individual Rho GTPases from RhoGDI. For example, our laboratory showed that diacylglycerol kinase zeta (DGKz), which converts diacylglycerol (DAG) to phosphatidic acid (PA), activates PAK1-mediated RhoGDI phosphorylation on Ser-101/174, causing selective Rac1 release and activation. Phosphorylation of RhoGDI on Ser-34 by PKCa has recently been demonstrated to selectively release RhoA, promoting RhoA activation. Here, I show DGKz is required for optimal RhoA activation and RhoGDI Ser-34 phosphorylation. Both were substantially reduced in DGKz-null fibroblasts and occurred independently of DGKz activity, but required a function DGKz PDZ-binding motif. In contrast, Rac1 activation required DGKz-derived PA, but not PDZ-interactions, indicating DGKz regulates these Rho GTPases by two distinct regulatory complexes. Interestingly, RhoA bound directly to the DGKz C1A domain, the same region known to bind Rac1. By direct interactions with RhoA and PKCa, DGKz was required for the efficient co-precipitation of these proteins, suggesting it is important to assemble a signalling complex that functions as a RhoA-specific RhoGDI dissociation complex. Consequently, cells lacking DGKz exhibited decreased RhoA signalling downstream and disrupted stress fibers. Moreover, DGKz loss resulted in decreased stress fiber formation following the expression of a constitutively active RhoA mutant, suggesting it is also important for RhoA function following activation. This is consistent with the ability of DGKz to bind both active and inactive RhoA conformations. Collectively, these findings suggest DGKz is central to two distinct Rho GTPase activation complexes, each having different requirements for DGKz activity and PDZ interactions, and might regulate the balance of Rac1 and RhoA activity during dynamic changes to the actin cytoskeleton.

Yes / Cells migrating over 2D substrates are required to polymerise actin at the leading edge to form lamellipodia protrusions and nascent adhesions to anchor the protrusion to the substrate. The major actin nucleator in lamellipodia formation is the Arp2/3 complex, which is activated by the WAVE regulatory complex (WRC). Using inducible Nckap1 floxed mouse embryonic fibroblasts (MEFs), we confirm that the WRC is required for lamellipodia formation, and importantly, for generating the retrograde flow of actin from the leading cell edge. The loss of NCKAP1 also affects cell spreading and focal adhesion dynamics. In the absence of lamellipodium, cells can become elongated and move with a single thin pseudopod, which appears devoid of N-WASP. This phenotype was more prevalent on collagen than fibronectin, where we observed an increase in migratory speed. Thus, 2D cell migration on collagen is less dependent on branched actin.

No description available.

Les travaux présentés se sont intéressés à la régulation des forces produites par le cytosquelette d'actine. Le rôle primordial joué par le microenvironnement a été au centre de nos investigations. L'étude de ces phénomènes a nécessité le développement de techniques innovantes. La première permet le contrôle en temps réel de la forme de la cellule. Elle utilise un laser UV pulsé pour modifier le microenvironnement adhésif de la cellule et contrôler les zones disponibles pour son étalement. La seconde est une amélioration d'une technique existante au sein du laboratoire. Il s'agit de produire des îlots de protéines d'adhésions, de forme contrôlée, sur un substrat déformable d'acrylamide. Ces supports permettent le contrôle de la taille de la cellule et de son organisation interne. En outre, l'élasticité de l'acrylamide permet la mesure des forces générées par la cellule. La dernière technique a combiné le patterning sur acrylamide avec l'ablation laser. Les forces produites au sein d'une structure particulière du cytosquelette ont ainsi pu être estimées. Deux grands mécanismes de régulation des forces ont pu être mis en évidence. L'utilisation de techniques de spectrométrie de masse, de mesure de forces et de biologie moléculaire a permis de mettre en évidence la coopération entre les différents types d'intégrines au niveau de l'adhésion cellulaire. Cette coopération permet un couplage entre l'architecture du cytosquelette et la quantité de moteurs moléculaires mettant en tension ces structures. Ces mécanismes sont primordiaux pour l'adaptation de la cellule à la rigidité de son environnement. Ce sont les structures d'actine qui produisent les forces qui seront transmises au niveau des adhésions. La corrélation entre la taille de ces structures et les forces générées est encore mal caractérisée. La relation entre taille des fibres de stress et répartition des forces au sein de la cellule a pu être étudiée et suggère que la force produite par une fibre de stress augmente avec sa longueur. Une étude systématique de la contractilité des cellules, sur des patterns de différentes tailles, a permis de montrer la relation entre la taille des fibres de stress et la force générée. Une relation biphasique a ainsi été mise en évidence. Quand la taille de la cellule augmente, la force générée au sein des fibres de stress commence par augmenter avant de diminuer au delà d'une longueur critique. Cette longueur correspond également à la taille maximale observée sur des cellules libres de s'étaler sans contraintes. Les résultats obtenus suggèrent que cette chute de force est liée à une augmentation excessive du ratio myosine/actine qui ne permet plus une production de force efficace. Le mécanisme pourrait faire intervenir le désassemblage des structures d'actine par la myosine ou la quantité insuffisante d'actine pour permettre un travail efficace des moteurs moléculaires. La rencontre de ces deux mécanismes permet de définir le champ des possibles pour la cellule en terme de contractilité. Le mécanisme de chute de forces observé n'a pas pu être expliqué à ce jour mais nous travaillons activement pour qu'il le soit dans les mois à venir. Ce phénomène aura sans doute un grand rôle à jouer dans l'intégrité mécanique des tissus et les phénomènes de migration. La chute de force au delà de la longueur critique permet en effet de déstabiliser les adhésions et pourrait être à l'origine de la rétraction de la cellule dans la migration ou du détachement d'une cellule de ces voisines dans le cas d'un tissu sous forte contraintes. Ce détachement protégerait ainsi la cellule d'un déchirement sous l'effet de forces trop importantes. / Our work has been focused on the regulation of the forces generated by the actin cytoskeleton. We have more precisely studied the role of the cellular microenvironment in this process. It was necessary to overcome some technical challenges to study these mechanisms. We developed two new techniques. The first one allows for the dynamic control of cell shape. A pulsed UV laser is used to modify the adhesive microenvironment around the cell and to create new area available for cell spreading. The second technique is an improvement of an existing technique from the laboratory. It consists in producing ECM protein islands on a elastic acrylamide substrate. This substrate provides the control of cell shape and internal organization. Plus, the elasticity of the substrate is compatible with traction forces measurements. The last technique combines acrylamide micropatterning and laser ablation of intracellular actin structures. Thus, the forces produced by a particular intracellular structure can be estimated. Two keys mechanisms of force regulation were shown. The use of mass spectrometry, traction force microscopy and molecular biology made it possible to study the interaction between different integrins in the adhesion complex. Cooperation was shown. It allows for the coupling between the architecture of the cytoskeleton and the amount of molecular motors in action. This process is necessary for the adaptation of cell forces to substrate stiffness. Actin structures are the one responsible for force production. This force can then be transmitted to the environment through adhesions.. The link between the length of actin fibers and the force produced was more precisely studied. The results showed a correlation between stress fibers length and the force generated inside it. This was true only above a certain critical value. After that, the force was rather decreasing with increasing fiber length. This critical length corresponds to the maximal length of cell axis on infinite 2D substrate. Our main hypothesis is that a too high myosin/actin ratio will block the proper force production/transmission within the fiber. Disassembly of actin by myosin or limited pool of actin are the two explanations we are currently following. The combination of these two-regulation process put brakes on force production by the cell. Above a certain length, the force produced is decreasing. This decreases in turn the strength of the adhesions anchored to these fibers. This will destabilize the adhesions and causes cell retraction The interplay between the regulation by the adhesion and the production of forces within the fiber set some limits on the level of forces produced by the cell. These processes are likely to be modified in a pathological context and can lead to tumor formation. They also protect the cell from being destroyed by stretching. If the length/stretch is too high, the cell will decrease its forces and detach from neighboring cells. This provide a system protecting the cell from being destroyed by massive deformations within the body

Cell contractility is mainly imagined as a force dipole-like interaction based on actin stress fibers that pull on cellular adhesion sites. Here, we present a different type of contractility based on isotropic contractions within the actomyosin cortex. Measuring mechanosensitive cortical contractility of suspended cells among various cell lines allowed us to exclude effects caused by stress fibers. We found that epithelial cells display a higher cortical tension than mesenchymal cells, directly contrasting to stress fiber-mediated contractility. These two types of contractility can even be used to distinguish epithelial from mesenchymal cells. These findings from a single cell level correlate to the rearrangement effects of actomyosin cortices within cells assembled in multicellular aggregates. Epithelial cells form a collective contractile actin cortex surrounding multicellular aggregates and further generate a high surface tension reminiscent of tissue boundaries. Hence, we suggest this intercellular structure as to be crucial for epithelial tissue integrity. In contrast, mesenchymal cells do not form collective actomyosin cortices reducing multicellular cohesion and enabling cell escape from the aggregates.

Η διδακτορική μου διατριβή εστιάζει στη λειτουργική ανάλυση του μοριακού μηχανισμού λειτουργίας της ιντεγκρινο-συνδεόμενης κινάσης (ILK) κατά την ανάπτυξη στη Drosophila. Έγινε διερεύνηση: α) της λειτουργικής συντήρησης της ILK κατά την εξέλιξη, β) του πιθανού ρόλου συγκεκριμένων αμινοξικών μοτίβων στον κυτταρικό εντοπισμό και τη λειτουργία της πρωτεΐνης και γ) των λειτουργικών ιδιοτήτων της ILK όταν είναι ομοιοπολικά συζευγμένη στην πλασματική μεμβράνη. Παράλληλα, αναζητήσαμε νέους λειτουργικούς ρόλους της ILK κατά την ανάπτυξη: α) σε άλλους ιστούς εκτός από το μυϊκό σύστημα και β) στην ωογένεση. Η ILK στο επίπεδο της αμινοξικής αλληλουχίας παρουσιάζει 60% ταυτότητα και 75% ομολογία με την αντίστοιχη πρωτεΐνη των θηλαστικών. Με βάση αυτή την ομολογία, ελέγξαμε την πιθανή φυλογενετική συντήρηση της λειτουργίας της ILK. Για το σκοπό αυτό κατασκευάστηκαν διαγονιδιακά στελέχη που περιείχαν την κωδική περιοχή της ILK του ανθρώπου (hILK) και του ποντικού (mILK) αντίστοιχα. Η ILK του ποντικού και του ανθρώπου ακολουθεί παραπλήσιο πρότυπο υποκυτταρικής κατανομής με την ενδογενή πρωτεΐνη στα μυϊκά κύτταρα Drosophila. Οι δυο ετερόλογες πρωτεΐνες υποκαθιστούν τη λειτουργία της ILK στη Drosophila. Ωστόσο, η ILK του ανθρώπου παρουσιάζει μειωμένη δυνατότητα πρόσδεσης με την parvin της Drosophila. Προκειμένου να διερευνήσουμε το μοριακό μηχανισμό ρύθμισης και δράσης της ILK κατά την ανάπτυξη, ελέγξαμε εάν η φωσφορυλίωση των αμινοξέων S176 και T180 υπεισέρχεται στη ρύθμιση της λειτουργίας της ILK. Σε πειράματα κυτταρικών σειρών έχει δειχθεί ότι στις θέσεις αυτές η φωσφορυλίωση ελέγχει την υποκυτταρική κατανομή της πρωτεΐνης στον πυρήνα. Ωστόσο, αποδείξαμε ότι η πιθανή φωσφορυλίωση στις ισχυρά συντηρημένες θέσεις S176 και T180 δεν είναι απαραίτητη για τον εντοπισμό της ILK στις μυοτενοντικές συνδέσεις και τη λειτουργία της ILK. Ένα άλλο αμινοξικό κατάλοιπο που είναι απαραίτητο για τον εντοπισμό της ILK στις εστιακές θέσεις προσκόλλησης είναι το F436, το οποίο εδράζεται στην τελευταία α έλικα του καρβοξυτελικού λοβού της περιοχής κινάσης. Ο υποκυτταρικός εντοπισμός της ILK και η λειτουργία της δεν επηρεάζονται από τη σημειακή μεταλλαγή F436Α σε αντίθεση με πειράματα σε κυτταρικά μοντέλα. Η σημειακή μεταλλαγή F436A αποδυναμώνει την ικανότητα αλληλεπίδρασης της ILK με την parvin. Εξετάσαμε αν η μεμβρανοδεσμευόμενη ILK μέσω παλμυτυλίωσης ή φαρνεσυλίωσης μπορεί να υποκαταστήσει την απουσία της ενδογενούς, καθώς και εάν μπορεί να προσελκύσει πρωτεΐνες του συνδεοσώματος ανεξάρτητα των ιντεγκρινών. Κατασκευάστηκαν δυο εναλλακτικές μορφές μεμβρανοδεσμευμένης ILK, οι GAP-ILK-GFP και ILK-GFP-HRAS εντοπίζονται με επιτυχία σταθερά στην πλασματική μεμβράνη των εμβρυϊκών μυϊκών κυττάρων. Επίσης, οι GAP-ILK-GFP και ILK-GFP-HRAS υποκαθιστούν πλήρως την ενδογενή ILK σε όλα τα αναπτυξιακά στάδια της ζωής της μύγας. Τέλος, η GAP-ILK-GFP συγκεντρώνει στις ΜΤΣ τα άλλα δυο μέλη του IPP συμπλόκου αλλά και την talin στις ΜΤΣ εμβρύων τελικού σταδίου τόσο αγρίου τύπου όσο και ομοζυγώτων για aPS2. Παράλληλα, μελετήσαμε , τόσο σε γενετικό όσο και σε μοριακό επίπεδο, το ρόλο της ILK στη μορφογένεση του ωοθυλακίου, την οργάνωση και ομοιόσταση του θυλακώδους επιθηλίου κατά τη διάρκεια της ωογένεσης στη Drosophila. Αποσιωπήσαμε την ilk κατασκευάζοντας γενετικά μωσαϊκά αλλά και ιστοειδικά διασωσμένα άτομα. Παρατηρήσαμε ότι η ILK είναι απαραίτητη για τη διαδικασία της ωογένεσης στη μύγα. Η απουσία της ILK προκαλεί διαταραχές στο σχηματισμό των διαθυλιακών μίσχων και ανωμαλία στο διαχωρισμό των νεοσχηματιζόμενων διαδοχικών ωοθυλακίων (σιαμαία ωοθυλάκια). Επίσης, τα πειράματά μας αποκάλυψαν ότι η ILK είναι απαραίτητη για την οργάνωση των ινιδίων ακτίνης κατά τα τελευταία στάδια της ωογένεσης και για την ομοιόσταση του κυτταροσκελετού ακτίνης κατά τον ακραιο-βασικό άξονα του κυττάρου. Ακόμη, η ILK είναι απαραίτητη για την οργάνωση και διατήρηση των βασο-πλευρικών κυτταρικών συνδέσεων στα θυλακιοκύτταρα, αλλά όχι των συνδέσεων ζώνης. Η απουσία της ILK διαταράσσει τον εντοπισμό των ιντεγκρινών στα άκρα των ινιδίων τάσης της ακτίνης στα θυλακιοκύτταρα των τελευταίων σταδίων. Επιπλέον, η ILK συμμετέχει στη ρύθμιση της δυναμικής της F-ακτίνης μειορρυθμίζοντας το Dia και αυξορρυθμίζοντας την profilin. H ILK εμπλέκεται στον έλεγχο της συσταλτότητας των ινιδίων ακτο-μυοσίνης στα θυλακιοκύτταρα των τελευταίων σταδίων, πιθανότατα μέσω της διαταραχής στον υποκυτταρικό εντοπισμό του RhoI, καθώς και μέσω της εκτοπικής συσσώρευσης της μυοσίνης (zipper). Τέλος, η ilk αλληλεπιδρά γενετικά με τη dpak στο επιθήλιο του ωοθυλακίου. Η ΙLK επηρεάζει τον εντοπισμό της dPAK στα θυλακιοκύτταρα τελευταίων σταδίων. Επίσης, η dPAK είναι απαραίτητη για τον εντοπισμό των ιντεγκρινών και της ILK στα άκρα των ινιδίων τάσης της ακτίνης. Ενώ, η απουσία της dpak, όπως και της ilk, διαταράσσoυν την οργάνωση και των ινιδίων τάσης της ακτίνης σε θυλακιοκύτταρα τελευταίων σταδίων. / My thesis is focused on the functional analysis of the molecular mechanism of the integrin-linked kinase (ILK) during development in Drosophila. We studied: a) the functional conservation of ILK in evolution, b) the possible role of specific amino acid motifs in the subcellular localization and function of ILK and c) the functional properties of ILK, when covalently bound to the plasma membrane. Furthermore, we sought new functional roles for ILK during development: a) in other tissues besides muscle system and b) in oogenesis. ILK protein sequence shares 60% identity and 75% similarity with the mammalian ILK. Based on these data, we tested the possible phylogenetic conservation of ILK function. For this purpose, we generated transgenic lines carrying the coding sequence of either human ILK (hILK) or mouse ILK (mILK). The mammalian ILK has localizes similarly to the endogenous protein, in the muscle cells of Drosophila. Both mammalian proteins can substitute for the ILK function in Drosophila. However, human ILK binds to Dparvin with reduced affinity compared to the fly ILK. In order to investigate the molecular mechanism through which ILK regulates and acts during development, we tested whether the phosphorylation on the amino acids S176 and T180 contributes to the regulation of ILK function. It has been shown, in cell culture models, that the phosphorylation on these sites controls the subcellular localization of the protein in the nucleus. However, we proved that the possible pgoshorylation of these highly conserved residues is dispensable for the ILK localization at the muscle attachment sites (MAS) as well as for the function of ILK. Another residue which is necessary to localise ILK at the focal adhesion sites is F436. It is located on the last a helix of the carboxyl-terminal lobe of the kinase-like domain. The subcellular localization and the ILK function are unaffected by the point mutation F436A, in contrast to the experimental data on cell culture models. The point mutation F436A affects the ability of ILK to bind to parvin. We examined, whether membrane-bound ILK, through palmytoylation or farnesylation, is able to substitute the absence of the endogenous ILK, if ii can recruit proteins of the adhesome, independently of integrins. We generated two alternative forms of membrane-bound ILK, GAP-ILK-GFP and ILK-GFP-HRAS, which both localize successfully at the plasma membrane of the embryonic muscle cells. Also, GAP-ILK-GFP and ILK-GFP-HRAS can substitute for the endogenous ILK throughout development. Moreover, GAP-ILK-GFP is able to recruit both PINCH and Parvin, as well as talin at the MAS, in both wild type and aPS2 mutant embryonic muscle cells. Furthermore, we studied, in genetic molecular level, the role of ILK in the morphogenesis of the egg chambers, the organization and the homeostasis during oogenesis in Drosophila. We used two experimental approaches in order to silence ilk: a) we generated genetic mosaics for ilk and b) we used conditionally rescued ilk-/- flies. We observed that ILK is indispensable for the process of oogenesis in the fly. Loss of ILK disrupts the stalk cell formation and the separation of the successive newly formed egg chambers (twin egg chambers). Also, our experiments revealed that ILK is essential for the organization of the actin stress fibers at the late developmental stages of oogenesis and for the homeostasis of the actin cytoskeleton along apico-basal axis of the cell. ILK is indispensable for the organization and the maintenance of the baso-lateral cell junctions in the follicle cells, but not for the adherens junctions. Loss of ILK disrupts the localization of integrins at the tips of the actin stress fibers of the follicle cells at late developmental stages. Moreover, ILK participates in the regulation of the F-actin dynamics by down-regulating Dia and up-regulating profilin. ILK is involved in the control of the contractility of the acto-myosin fibers in the follicle cells at late developmental stages, probably by affecting the subcellular localization of Rho1, and causing ectopic accumulation of myosin (zipper). Finally, ilk interacts genetically with dpak in the follicular epithelium. ILK affects dPAK localization in the follicle cells at late developmental stages. Furthermore, dPAK is essential for the localization of both integrins and ILK at the tips of actin stress fibers. Loss of dpak, similarly to ilk, disrupts the organization of actin stress fibers in follicle cells at late developmental stages.