α-actinin-4 is a non-muscle isoform of α-actinin that belongs to the spectrin superfamily. It comprises three functional regions: an N-terminal actin-binding region that consists of two calponin homology (CH) domains, a central region that consists of four copies of the spectrin-like repeat domain and a C-terminal calmodulin-like domain that is predicted to bind Ca²⁺. α-actinin-4 is organised as an antiparallel homodimer formed by the interaction of four spectrin-like repeats between the two monomers, giving a rod-like shape, with actin binding regions at both ends. α-actinin-4 is an abundant actin-bundling protein, which provides a direct link between actin filaments and integrins, and is believed to play an important role in stabilising cell shape and adhesion and regulating cell migration. It also acts as a tumor suppressor and influences the metastatic potential and invasiveness in human cancers. A cluster of three actin binding motifs have been identified in the CH domains (2X CH) from other members of the spectrin superfamily, utrophin and dystrophin. Two of them reside in the CH1 domain and the third resides in the first α-helix of the CH2 domain. In addition, a PIP2 binding site has been mapped on a region adjacent to actin-binding site-3. These observations imply the F-actin binding activity would be regulated by phosphoinositides. Five mutations of α-actinin-4, K122N, an alternative splice variant, K255E, T259I and S263P, have been reported to be involved in three human diseases, non-small lung cancer (NSCLC), small cell lung cancer (SCLC) and focal segmental glomerulosclerosis (FSGS). The mutation site within these mutants is located on the actin binding region. Therefore, the actin binding region is presumed to be associated with the progression of human disease. The aims of this thesis focused on the regulation of the F-actin binding activity of α-actinin-4 by phosphoinositides (PIP2 and PIP3), the calmodulin-like domain and Ca²⁺ , determination of the three-dimensional structure of the CH2 domain in solution and identification of the phosphoinositide binding site on the CH2 domain. In order to investigate the F-actin binding activity quantitatively, a novel in vitro F-actin binding assay (solid phase) was established to replace the semi-quantitative actin bundling assay. Using this novel solid phase F-actin binding assay, Ca²⁺ was shown to enhance the F-actin binding activity of α-actinin-4 in a concentration-dependent manner. The presence of 10 mM Ca²⁺ results in a two-fold increase in the F-actin binding activity. Both PIP2 and PIP3 inhibited the F-actin-binding activity of α-actinin-4 in a concentration-dependent manner with an approximate IC₅₀ of 75 and 45 μM, respectively. In order to characterise how phosphoinositides regulated the F-actin binding activity of α-actinin-4, the solution structure of α-actinin-4 CH2 domain was determined and the phosphoinositide binding residues within the CH2 domain were identified using NMR spectroscopy. The solution structure of α-actinin-4 CH2 domain contained six α-helices and was similar to that of other spectrin superfamily members. The strategy used in identification of the phosphoinositide binding site was an NMR-based 2D ¹H-¹⁵N HSQC ligand titration assay to replace the traditional semi-quantitative protein-lipid overlay assay. Using the NMR-based ligand titration assay, the recognition site for the inositol head group resides in residues Trp 172, Tyr 265 and His 266 and the binding region of acyl chains resides in the first α-helix structure which is one of the putative F-actin binding sites. In order to examine the interaction of phosphoinositides with this site, Y265A and H266E mutants of α-actinin-4 CH2 domain were generated using site-directed mutagenesis and verified the interaction with phosphoinositides and the inositol head group using an NMR-based ligand titration assay. These results confirmed the phosphoinositide binding site on the CH2 domain and residues, Tyr 265 and His 266, are critical for interacting with phosphoinositides. Wildtype and mutants (Y265A and H266E) of α-actinin-4 were expressed in mammalian cells as EGFP-fusion proteins. Wildtype α-actinin-4 was shown to be co-localised with focal adhesions and actin stress fibres. However, Y265A and H266E mutants of α-actinin-4 were co-localised with actin stress fibres but poorly co-localised with focal adhesions. Moreover, both Y265A and H266E mutants of α-actinin-4 were co-localised with actin in the cytoplasm rather than localised along the cell membrane after EGF stimulation for 30 minutes. These results suggested that PIP2 assists the co-localisation of α-actinin-4 with focal adhesions. Taken together, the results described in this thesis concluded that Ca²⁺ enhanced the F-actin binding activity of α-actinin-4 in vitro. However, phosphoinositides (PIP2 or PIP3) inhibited the F-actin binding activity in vitro. Moreover, the results described in this thesis provided a phosphoinositide binding site on α-actinin-4 CH2 domain. Binding to PIP2 is important to the localisation of α-actinin-4 in focal adhesions. / Thesis (Ph.D.) -- University of Adelaide, School of Molecular and Biomedical Science, 2008

<p>alpha-actinin is a ubiquitous protein found in most eukaryotic organisms. The ability to form dimers allows alpha-actinin to cross-link actin in different structures. In muscle cells alpha-actinin is found at the Z-disk of sarcomeres. In non-muscle cells alpha-actinin is found in zonula adherens or focal adhesion sites where it can bind actin to the plasma membrane.</p><p>alpha-actinin is the shortest member of the spectrin superfamily of proteins which also includes spectrin, dystrophin and utrophin. Several hypotheses suggest that alpha-actinin is the ancestor of this superfamily.</p><p>The structure of alpha-actinin in higher organisms has been well characterized consisting of three main domains: an N-terminal actin-binding domain with two calponin homology domains, a central rod domain with four spectrin repeats and a C-terminal calcium-binding domain. Data mining of genomes from diverse organisms has made possible the discovery of new and atypical alpha-actinin isoforms that have not been characterized yet.</p><p>Invertebrates contain a single alpha-actinin isoform, whereas most of the vertebrates contain four. These four isoforms can be broadly classified in two groups, muscle isoforms and non-muscle isoforms. Muscle isoforms bind actin in a calcium independent manner whereas non-muscle isoforms bind actin in a calcium-dependent manner.</p><p>Some of the protozoa and fungi isoforms are atypical in that they contain fewer spectrin repeats in the rod domain. We have purified and characterized two ancestral alpha-actinins from the parasite Entamoeba histolytica. Our results show that despite the shorter rod domain they conserve the most important functions of modern alpha-actinin such as actin-bundling formation and calcium-binding regulation. Therefore it is suggested that they are genuine alpha-actinins.</p><p>The phylogenetic tree of alpha-actinin shows that the four different alpha-actinin isoforms appeared after the vertebrate-invertebrate split as a result of two rounds of genome duplication. The atypical alpha-actinin isoforms are placed as the most divergent isoforms suggesting that they are ancestral isoforms. We also propose that the most ancestral alpha-actinin contained a single repeat in its rod domain. After a first intragene duplication alpha-actinin with two spectrin repeats were created and a second intragene duplication gave rise to modern alpha-actinins with four spectrin repeats.</p>

alpha-actinin is a ubiquitous protein found in most eukaryotic organisms. The ability to form dimers allows alpha-actinin to cross-link actin in different structures. In muscle cells alpha-actinin is found at the Z-disk of sarcomeres. In non-muscle cells alpha-actinin is found in zonula adherens or focal adhesion sites where it can bind actin to the plasma membrane. alpha-actinin is the shortest member of the spectrin superfamily of proteins which also includes spectrin, dystrophin and utrophin. Several hypotheses suggest that alpha-actinin is the ancestor of this superfamily. The structure of alpha-actinin in higher organisms has been well characterized consisting of three main domains: an N-terminal actin-binding domain with two calponin homology domains, a central rod domain with four spectrin repeats and a C-terminal calcium-binding domain. Data mining of genomes from diverse organisms has made possible the discovery of new and atypical alpha-actinin isoforms that have not been characterized yet. Invertebrates contain a single alpha-actinin isoform, whereas most of the vertebrates contain four. These four isoforms can be broadly classified in two groups, muscle isoforms and non-muscle isoforms. Muscle isoforms bind actin in a calcium independent manner whereas non-muscle isoforms bind actin in a calcium-dependent manner. Some of the protozoa and fungi isoforms are atypical in that they contain fewer spectrin repeats in the rod domain. We have purified and characterized two ancestral alpha-actinins from the parasite Entamoeba histolytica. Our results show that despite the shorter rod domain they conserve the most important functions of modern alpha-actinin such as actin-bundling formation and calcium-binding regulation. Therefore it is suggested that they are genuine alpha-actinins. The phylogenetic tree of alpha-actinin shows that the four different alpha-actinin isoforms appeared after the vertebrate-invertebrate split as a result of two rounds of genome duplication. The atypical alpha-actinin isoforms are placed as the most divergent isoforms suggesting that they are ancestral isoforms. We also propose that the most ancestral alpha-actinin contained a single repeat in its rod domain. After a first intragene duplication alpha-actinin with two spectrin repeats were created and a second intragene duplication gave rise to modern alpha-actinins with four spectrin repeats.

Ubiquitin C-terminal hydrolase L1 is a deubiquitinating enzyme that salvages ubiquitin from substrates and maintains intracellular ubiquitin pools. While the role of ubiquitin C-terminal hydrolase L1 is well characterized in neurons, there is an increasing scope of evidence to suggest that ubiquitin C-terminal hydrolase L1 also plays a role in renal function and glomerular disease, however, its specific role in these settings remains incompletely elucidated. In the present thesis we explored the role of ubiquitin C-terminal hydrolase L1 in a mouse model of glomerular disease, ACTN4-associated focal segmental glomerulosclerosis and the role of ubiquitin C-terminal hydrolase L1 in renal function. Deletion of ubiquitin C-terminal hydrolase L1 in a mouse model of ACTN4-associated focal segmental glomerulosclerosis significantly improved indices of podocyte injury, a likely result of ubiquitin pool attenuation and sustained α-actinin-4 levels. However, global ablation of ubiquitin C-terminal hydrolase L1 in mice led to altered renal hemodynamics, namely glomerular hyperfiltration, most likely attributed to nerve dysfunction and loss of arterial resistance. Finally, mice lacking ubiquitin C-terminal hydrolase L1 exhibited perturbations in phosphate homeostasis as these showed evidence of hyperphosphatemia and phosphaturia, indicating altered renal phosphate balance. Altogether, these data show that while ubiquitin C-terminal hydrolase L1 plays a maladaptive role in glomerular disease, it also functions as a crucial regulator of renal hemodynamics and renal phosphate handling, suggesting that it may have distinct functions in diseased and non-diseased kidneys.