Final Activity Report Summary - CYTOSKELETAL MODULES (The molecular basis for actin cytoskeleton regulation by functional protein modules) Work over the past four years aimed at investigating the principles of modular proteins that regulate cytoskeletal organisation with a particular emphasis on those factors that regulate the formation of actin polymerisation-dependent invasion-promoting cell adhesions. Activities also addressed the general assembly principles underlying the functional plasticity of protein modules in cytoskeleton and signalling proteins, and aimed to correlate these findings with the molecular mechanisms driving stability and dynamic modulation of the actin cytoskeleton. Functional diversity of proteins is driven by subtle modulations of sequence elements dispensable for the maintenance of a module's three dimensional structure. Thus, a large number of functionally diverse molecules can be generated from a small number of stable structural folds. Based on the concept of compositional semantics the functions in a protein can be generated in a rule-based fashion. We used a selection of protein modules (CH domains, CLIK modules, and LIM domains) that play key roles in the organization of the actin cytoskeleton, the maintenance of smooth muscle tone and tension in fibroblast cells, and the regulation of actin-myosin interactions, as models for our studies. The individual overlapping projects aimed at gaining insight into the regulation of cytoskeleton stability and dynamics by modular protein domains. We could show that a major Rho GTPase inactivator, p190RhoGAP, is recruited to podosomes by the multiple SH3 domain-containing adapter molecule Tks5/FISH, and that GAP-defective as well as mistargeted p190 mutants abrogate podosome formation. We also showed that the giant multi-domain cytolinker protein plectin accumulates in the rings that surround podosomes and regulates growth and maturation of podosomes by reducing focal adhesion and stress fiber turnover. Pharmacological inhibition of cellular contractility leads to an aberrant localisation of podosomes and induces failure of plectin to surround the outer perimeter of these invasive adhesions. The actin-binding domains of many proteins consist of a canonical type1/type2 arrangement of the structurally conserved calponin homology (CH) domain. Using the actin-binding domain (ABD) of a-actinin-1 as a scaffold we have generated synthetic actin binding domains by altering position and composition of the CH domains. We could show that the presence of two CH domains in an ABD is not sufficient for actin binding; CH domains in an inverted position, however, function normally, demonstrating that the dynamics and specificity of ABDs requires both the filament binding properties of the type1, and regulation by type2 CH domains, but is independent of their position. The LIM domain-containing protein Cysteine-Rich Protein 2 (CRP2) activates the transcriptional machinery via GATA4 / GATA6 and SRF-dependent circuits. Over-expression of CRP2 stabilises the actin cytoskeleton and CRP2-decorated actin fibers display reduced turnover and podosome formation. We have now established a dual role for CRP2 as an actin binding protein and as a unique mechanosensor that responds to alterations in the polymerisation status of the cytoskeleton. We have shown that CRP2 is sensitive to changes in the polymerization status of the actin cytoskeleton and is rapidly transported from the cytoplasm to the nucleus in response to alterations in the polymerisation status of the actin cytoskeleton, and to actomyosin-dependent contractility. We also identified Lys91 as a potential site for SUMOylation (a post-translational protein modification) that is situated between the two zinc-finger domains of CRP2, which influences the ability of CRP2 to shuttle between the cytoplasmic and nuclear compartment.
