Actin filaments are integral to a large number of cellular structures. Due to the widespread use of actin, cells must rigorously control when and where new filaments assemble. Formin proteins are expressed in all eukaryotes and provide the potential to regulate actin dynamics. They do so by acting as actin assembly factors that accelerate filament nucleation, then remain on the elongating barbed end and modulate filament elongation. The large number of mammalian formins (15 formin genes) allows them to be diverse in their effects on actin and actin-based structures. One such structure, the filopodium, is of particular interest since multiple formins have been associated with its assembly. However, the precise mechanism that governs formin-mediated filopodial assembly remains to be fully understood.;In this thesis I characterize the cellular and biochemical properties of the mammalian formin FMNL3. A constitutively active FMNL3 construct, comprised of its FH1 and FH2 domains, induces filopodial formation in Jurkat T lymphocytes and U2OS osteosarcoma cells and also enriches at the tips of filopodia. The FH2 domain is the defining feature of all formins and is capable of directly interacting with both actin monomers and filaments, while the FH1 domain acts as a binding site for profilin. Mutations in the FH2 of FMNL3 that abolish barbed end binding also abolish its ability to induce filopodia. The FH1 domain, though required for filopodia assembly, does not supply specificity, since FMNL1's FH1 can substitute. At present, it is unclear whether an additional actin nucleator, Arp2/3 complex, is required for FMNL3-mediated filopodial assembly, since the Arp2/3 inhibitor CK666 inhibits assembly under some conditions but not others.;Biochemical analysis of a purified FMNL3 FH1-FH2 construct indicates that it is a poor actin nucleator. However, addition of FMNL3's 74 amino acid C terminus to the FH1-FH2 increases nucleation ability dramatically. Subsequent analysis revealed an actin binding site within the C terminus which resembling a WH2 domain. This binding site is capable of binding the barbed end of both actin monomers and actin filaments independently of the FH2 domain. Mutational analysis has identified key residues in the WH2-like domain responsible for actin binding and has led to a model describing FMNL3-mediated actin nucleation. In a collaborative effort with Jon Kull and Morgan Thompson in the Chemistry Department at Dartmouth College, the structure of a complex between FMNL3's FH2 dimer and two actin monomers has been solved to a 3.4 A&
展开▼
