@phdthesis{Offner2017,
  author    = {Offner, Kristin},
  title     = {SH3-mediated protein interactions of Mena and VASP},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-154481},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2017},
  abstract  = {Regulation of actin cytoskeletal turnover is necessary to coordinate cell movement and cell adhesion. Proteins of the Enabled/vasodilator-stimulated phosphoprotein (Ena/VASP) family are important mediators in cytoskeleton control, linking cyclic nucleotide signaling pathways to actin assembly. In mammals, the Ena/VASP family consists of mammalian Enabled (Mena), VASP, and Ena-VASP-like (EVL). The family members share a tripartite domain organization, consisting of an N-terminal Ena/VASP homology 1 (EVH1) domain, a central proline-rich region (PRR), and a C-terminal EVH2 domain. The EVH1 domain mediates binding to the focal adhesion proteins vinculin and zyxin, the PRR interacts with the actin-binding protein profilin and with Src homology 3 (SH3) domains, and the EVH2 domain mediates tetramerization and actin binding. Endothelial cells line vessel walls and form a semipermeable barrier between blood and the underlying tissue. Endothelial barrier function depends on the integrity of cell-cell junctions and defective sealing of cell-cell contacts results in vascular leakage and edema formation. In a previous study, we could identify a novel interaction of the PRR of VASP with αII-spectrin. VASP-targeting to endothelial cell-cell contacts by interaction with the αII-spectrin SH3 domain is sufficient to initiate perijunctional actin filament assembly, which in turn stabilizes cell-cell contacts and decreases endothelial permeability. Conversely, barrier function of VASP-deficient endothelial cells and microvessels of VASP- null mice is defective, demonstrating that αII-spectrin/VASP complexes regulate endothelial barrier function in vivo. The aim of the present study was to characterize the structural aspects of the binding of Ena/VASP proteins to αII-spectrin in more detail. These data are highly relevant to understand the cardiovascular function of VASP and its subcellular targeting. In the present study, the following points were experimentally addressed: 1. Comparison of the interaction between αII-spectrin and Mena, VASP, or EVL In contrast to the highly conserved EVH1/EVH2 domains, the PRR is the most divergent part within the Ena/VASP proteins and may differ in binding modes and mechanisms of regulation. More specifically, VASP contains a triple GP5 motif, whereas EVL and Mena contain one or more GP6 motifs or even longer proline stretches. In the present study, we used peptide scans and competitive αII-spectrin SH3 pull-down assays with the recombinant Mena, VASP, and VASP mutants to investigate the relative binding efficiency. Our results indicate that binding of the αII-spectrin SH3 domain to GP6 motifs is superior to GP5 motifs, giving a rationale for a stronger interaction of αII-spectrin with EVL and Mena than with VASP. 2. Interaction of SH3i with Ena/VASP proteins In the mammalian heart, an αII-spectrin splice variant exists (SH3i), which contains a 20 amino acid insertion C-terminal to the SH3 domain. We used GST-fusion proteins of αII-spectrin, comprising the SH3 domain with or without the alternatively spliced amino acids, to pull-down recombinant Mena, VASP or VASP mutants. The results demonstrate a substantially increased binding of the C-terminal extended SH3 domain as compared to the general αII-spectrin isoform without the 20 amino acid insertion. These findings were also confirmed in pull-down experiments with heart lysates and purified Mena from heart muscle. The increased binding was not due to an alternative, SH3-independent binding interface because a pointmutation of the SH3 domain (W1004R) in the alternatively spliced αII-spectrin isoform completely abrogated the interaction. To analyze the interaction of SH3i and Ena/VASP proteins in living cells, we expressed the extended SH3 domain as GFP fusion proteins in endothelial cells. Here, we observed an extensive co-localization with Mena and VASP at the leading edge of lamellipodia confirming the in vivo relevance of the interaction with potential impact on cell migration and angiogenesis. 3. Binding affinity and influence of the Ena/VASP tetramerization domain We also determined the binding affinity of the general and the alternatively spliced αII-spectrin SH3 with Ena/VASP proteins by isothermal titration calorimetry (ITC) using a peptide from the PRR of Mena (collaboration with Dr. Stephan Feller, University of Oxford). Surprisingly, the binding affinity of the general SH3 domain was low (~900 μM) as compared to other SH3 domain- mediated interactions, which commonly display binding constants in the low micromolar range. Furthermore and in contrast to the pull-down assays, we could not detect an increased binding affinity of the C-terminally extended SH3 domain. This could be either explained by the existence of a third protein, which "bridges" the Mena/αII-spectrin complex in the pull-down assays, or, more likely, by the small size of the Mena peptide, which lacks major parts of the Mena protein, including the tetramerization domain. Indeed, it has been previously shown that the tetramerization of Ena is crucial for the interaction with the Abl- SH3 domain, although no SH3 binding sites are found in the tetramerization domain. To address this point experimentally, we used a VASP mutant that lacks the tetramerization domain in pull-down assays. Neither the general nor the alternatively spliced SH3 domain bound to the monomeric VASP, demonstrating the crucial (indirect) impact of Ena/VASP tetramerization on the interaction with αII-spectrin. In summary, we conclude that the αII-spectrin SH3 domain binds to the proline- rich region of all Ena/VASP proteins. However, binding to EVL and Mena, which both possess one or more GP6 motifs, is substantially more efficient than VASP, which only contains GP5 motifs. The C-terminally extended SH3 domain, which is present in the αII-spectrin splice variant SH3i, binds stronger to the Ena/VASP proteins than the general isoform and expression of the isolated domain is sufficient for co-localization with Ena/VASP in living endothelial cells. Finally, the tetramerization of the Ena/VASP proteins is indispensable for the interaction with either isoform of αII-spectrin.},
  language  = {en}
}
@phdthesis{Fischer2010,
  author    = {Fischer, Andreas},
  title     = {The Role of Protein-Protein Interactions in the Activation Cycle of RAF Kinases},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-48139},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2010},
  abstract  = {Members of the RAF protein kinase family are key regulators of diverse cellular processes. The need for isoform-specific regulation is reflected by the fact that all RAFs not only display a different degree of activity but also perform isoform-specific functions at diverse cellular compartments. Protein-protein-interactions and phosphorylation events are essential for the signal propagation along the Ras-RAF-MEK-ERK cascade. More than 40 interaction partners of RAF kinases have been described so far. Two of the most important regulators of RAF activity, namely Ras and 14-3-3 proteins, are subject of this work. So far, coupling of RAF with its upstream modulator protein Ras has only been investigated using truncated versions of RAF and regardless of the lipidation status of Ras. We quantitatively analyzed the binding properties of full-length B- and C-RAF to farnesylated H-Ras in presence and absence of membrane lipids. While the isolated Ras-binding domain of RAF exhibit a high binding affinity to both, farnesylated and nonfarnesylated H-Ras, the full-length RAF kinases demonstrate crucial differences in their affinity to Ras. In contrast to C-RAF that requires carboxyterminal farnesylated H-Ras for interaction at the plasma membrane, B-RAF also binds to nonfarnesylated H-Ras in the cytosol. For identification of the potential farnesyl binding site we used several fragments of the regulatory domain of C-RAF and found that the binding of farnesylated H-Ras is considerably increased in the presence of the cysteine-rich domain of RAF. In B-RAF a sequence of 98 amino acids at the extreme N terminus enables binding of Ras independent of its farnesylation status. The deletion of this region altered Ras binding as well as kinase properties of B-RAF to resemble C-RAF. Immunofluorescence studies in mammalian cells revealed essential differences between B- and C-RAF regarding the colocalization with Ras. In conclusion, our data suggest that that B-RAF, in contrast to C-RAF, is also accessible for nonfarnesylated Ras in the cytosolic environment due to its prolonged N terminus. Therefore, the activation of B-RAF may take place both at the plasma membrane and in the cytosolic environment. Furthermore, the interaction of RAF isoforms with Ras at different subcellular sites may also be governed by the complex formation with 14-3-3 proteins. 14-3-3 adapter proteins play a crucial role in the activation of RAF kinases, but so far no information about the selectivity of the seven mammalian isoforms concerning RAF association and activation is available. We analyzed the composition of in vivo RAF/14-3-3 complexes isolated from mammalian cells with mass spectrometry and found that B-RAF associates with a greater variety of 14-3-3 proteins than C- and A-RAF. In vitro binding assays with purified proteins supported this observation since B-RAF showed highest affinity to all seven 14-3-3 isoforms, whereas C-RAF exhibited reduced affinity to some and A-RAF did not bind to the 14-3-3 isoforms epsilon, sigma, and tau. To further examine this isoform specificity we addressed the question of whether both homo- and heterodimeric forms of 14-3-3 proteins participate in RAF signaling. By deleting one of the two 14-3-3 isoforms in Saccharomyces cerevisiae we were able to show that homodimeric 14-3-3 proteins are sufficient for functional activation of B- and C-RAF. In this context, the diverging effect of the internal, inhibiting and the activating C-terminal 14-3-3 binding domain in RAF could be demonstrated. Furthermore, we unveil that prohibitin stimulates C-RAF activity by interfering with 14-3-3 at the internal binding site. This region of C-RAF is also target of phosphorylation as part of a negative feedback loop. Using tandem MS we were able to identify so far unknown phosphorylation sites at serines 296 and 301. Phosphorylation of these sites in vivo, mediated by activated ERK, leads to inhibition of C-RAF kinase activity. The relationship of prohibitin interference with 14-3-3 binding and phosphorylation of adjacent sites has to be further elucidated. Taken together, our results provide important new information on the isoform-specific regulation of RAF kinases by differential interaction with Ras and 14-3-3 proteins and shed more light on the complex mechanism of RAF kinase activation.},
  subject      = {Signaltransduktion},
  language  = {en}
}