γ-Tubulin is conserved and essential for microtubule nucleation from lower eukaryotes like unicellular fungi to higher eukaryotes like humans. γ-Tubulin together with other proteins forms complexes and functions in microtubule nucleation and organization. Although these γ-tubulin complex proteins (GCPs) have been identified in several model organisms, the functions of individual GCPs remain unclear and contradictory data from studies using different model organisms have prevented the establishment of a coherent model. Although the existence of GCPs in the filamentous fungus, Aspergillus nidulans, was suggested by early biochemical experiments (Akashi et al., 1997), these proteins remain unidentified and their functions remain undetermined. In my dissertation research, I have identified and deleted each of the genes that encode these proteins, created GFP and mCherry fusions of each of them, and observed them in vivo using spinning disk confocal microscopy. Briefly, I have found that different GCPs play different roles: two are required for the existence of the γ-tubulin complex in cells and are thus essential for formation of the mitotic apparatus and for cell reproduction. Three others are not essential, but play a role in the assembly of large γ-tubulin complexes and have minor functions in the fidelity of chromosomal segregation. We were also able to establish a hierarchy of binding of GCPs to the spindle pole body and, based on these findings, we were able to propose a model for the structure of the γ-tubulin complexes.
To understand the microtubule-nucleation and non-nucleation functions of γ-tubulin, our lab created a series of conditionally lethal γ-tubulin mutants. Extensive analysis of one mutant, mipAD159, revealed that the coordination of late mitotic events is disrupted even though mitotic spindles assemble with normal kinetics (Prigozhina et al., 2004). Further investigations revealed that this allele caused a nuclear autonomous failure of inactivation of the anaphase promoting complex/cyclosome (APC/C), which caused constitutive destruction of mitotic regulatory proteins (Nayak et al., 2010). As degradation of proteins targeted by the APC/C is carried out by the proteasome, we were interested in determining if γ-tubulin mutations affect the proteasome. Surprisingly, time-lapse imaging of a γ-tubulin mutant and γ-tubulin deletants as well as deletants of two essential GCPs revealed that the proteasome, which is highly enriched in the nucleus, separates physically from the chromosomes in mitosis. This resulted in the formation of nucleus-like structures containing little or no chromatin but often having a nucleolus. This implies that the proteasome binds to, or itself forms, a nuclear matrix and our high resolution multi-dimensional optical reconstructions of the nuclei support this notion. In an attempt to determine the nature of the nuclear matrix by using a candidate approach, I found that the nuclear localization of the proteasome does not rely on the Mlp1 (the orthologue of Megator) scaffold but does depend on the presence of the A. nidulans homolog of Schizosaccharomyces pombe Cut8 (An-Cut8). The lack of proteosomes in mitotic nuclei of An-cut8 deletants, however, does not cause a delay in cyclin B degradation and obvious defects in mitotic progression. My data do not seem to support the hypothesis that Cut8 as an anchor for the nuclear proteasome as has been proposed in S. pombe, instead, they appear more consistent with the notion that Cut8 is involved in the nuclear transport of the proteasome or proteasomal subunits.