Characterizing the role of Myosin VI at E-cadherin cell-cell adhesions

Epithelial cadherin (E-cad) cell-cell adhesion junctions play an important role in tissue organization and maintenance of tissue integrity. This is vital for many physiological processes such as morphogenesis during embryonal development, wound healing or tissue turnover (Takeichi, 1991). Disruption of cell-cell adhesion function not only has a negative impact on these essential physiological processes but also contributes to tumorigenesis and can lead to tumor progression into metastasis and invasion (Derksen et al., 2006). Therefore a detailed knowledge of the mechanisms of the formation, maintenance and disruption of E-cadherin adhesion junctions makes a vital contribution both towards the fields of molecular cell biology and cancer research. E-cad is a trans-membrane protein whose ectodomain engages in homophilic ligation and other complex adhesion mechanisms with cadherins on adjacent cells and thus contributes to cell-cell adhesion (Leckband and Prakasam, 2006). However its importance is not only as a structural protein, since it is also known to mediate dynamic processes such as cell polarity formation (Iden et al., 2006), and is able to modulate signaling at cell-cell adhesions (McLachlan and Yap, 2007). Its cytoplasmic tail plays an important part in the communication between adhesion junctions and the cytoskeleton (Goodwin and Yap, 2004; Mege et al., 2006). However the precise mechanism of this interaction is still subject to investigation. Myosin VI (MyoVI) is an unconventional actin-binding motor protein (Wells et al., 1999), which serves a multitude of important roles in the cell, both during development and in the adult organism. We are explicitly interested in the MyoVI pool that interacts with E-cad at cell-cell adhesions. Previous work showed that MyoVI engages physically with the E-cad/-catenin complex and plays an important role during the maturation and stabilization of newly forming contacts (Maddugoda et al., 2007). The physical properties of MyoVI imply that it could act as an anchor protein at adhesion junctions and thus contributes to junction stabilization (Altman et al., 2004; Oguchi et al., 2008). However to date very little is known about the precise mechanisms of MyoVI function or its regulation at E-cad adhesion junctions. The first part of this thesis is dedicated to the investigation of the biochemical properties of the interaction between E-cad and MyoVI. Direct binding experiments of recombinant proteins expressed in Escherichia coli (E. coli), revealed that E-cad and MyoVI interact via direct binding of their tail domains. Truncation analysis of the E-cad tail domain revealed that more than one binding site for MyoVI is likely to exist and that one of the binding sites is located in the membrane proximal 28 amino acids of the E-cad cytoplasmic tail. These results support the hypothesis that MyoVI could act as an anchor protein that is able to directly link the junctional E-cadherin/catenin complex to the actin cytoskeleton. By establishing that E-cad and MyoVI bind directly this study provides a novel insight into the organization of adherens junctions and a valuable basis for future mechanistic investigations. The second part of this thesis focuses on novel findings regarding the regulation and function of junctional MyoVI in Hepatocyte growth factor (HGF) treatment. HGF and its receptor, the tyrosine kinase c-met (Met), are molecular factors involved in Epithelial to mesenchymal transition (EMT), a well-investigated process in which cell-cell adhesions are disrupted and the cells develop a migratory phenotype. The process of EMT is important in embryonic development (Birchmeier and Gherardi, 1998), but also employed in tumor invasion and metastasis (Thiery, 2002; Vergara et al., 2010), which makes HGF induced Met signaling also an important target for cancer drug development (Cecchi et al., 2010; Sattler and Salgia, 2009). Upon HGF treatment I observed that MyoVI was lost from the zonula adherens (ZA) within 15 min. During the same time frame I also saw a dramatic re-arrangement of the apical F-actin cytoskeleton and a change in E-cad morphology. By means of over-expression of a GFP-tagged porcine MyoVI construct I could show that the HGF induced loss of MyoVI from the junctions was causal for the changes in F-actin organization. Fluorescence recovery after photobleaching (FRAP) and studies of a photoactivatable G-actin construct showed no difference in actin turnover dynamics in HGF treated versus untreated cells. Live cell studies however revealed a loss of anchorage of actin filaments at the ZA in HGF treatment, which could be counter-acted by over-expression of GFP-tagged MyoVI. The biochemical interaction between MyoVI and E-cad was reduced upon HGF treatment, which implied that the disruption of their interaction might be causal for the loss of actin filament retention at the ZA. Finally I could show that calcium signaling plays a crucial role in the HGF-induced regulation of junctional MyoVI and actin. Thus this study identifies MyoVI as a novel target of HGF signalling, which regulates actin filament retention at the ZA