Functional characterization of genes that control structure and dynamics of centrosomes in Drosophila

The general aim of my PhD thesis work was the characterization of Drosophila male meiosis, a widely used model system for genetic and molecular dissection of cell division. I first asked why the meiotic spindle of Drosophila males is much larger then the female meiotic spindle or the mitotic spindles of various cell types. I hypothesized that such large size reflects the availability of a large amount of tubulin stored in spermatocytes to be used postmeiotically for sperm tail assembly. To test this hypothesis, I examined male meiosis in 6 Drosophila species with dramatically different sperm flagella, ranging in length from 0.3 mm in D. persimilis to 58.3 mm in D. bifurca. I found that males of different species exhibit striking variations in meiotic spindle size, which positively correlate with sperm length, with D. bifurca showing a 30-fold larger spindle than D. persimilis. This suggests that primary spermatocytes of Drosophila species manufacture and store amounts of tubulin that are proportional to the length of the sperm tail and use these tubulin pools for spindle assembly. My findings also highlight an unsuspected plasticity of the meiotic spindle and suggest that the fidelity of chromosome segregation is largely independent of spindle size (Lattao et al. 2012, in press). I also performed a screen aimed at the identification of new D. melanogaster genes controlling centrosome structure and/or behavior. To facilitate this screen I developed two novel Tubby-marked balancer chromosomes that allow unambiguous recognition of mutant larvae from their heterozygous non-mutant siblings (Lattao et al., 2011). My screen led to the isolation of two mutants with abnormal numbers of centrosomes: molti asters (most) and fragile centrioles (fract). I characterized the meiotic phenotypes of these mutants and cloned the genes they specify. most encodes 749 aa transcription factor of the TFIID superfamily, with homology with the TBP-related factors. Mutations in most cause an arrest in the first meiotic prometaphase/metaphase and often exhibit multiple centrosomes in the aberrant meiotic figures. Thus, most belongs to a class of genes collectively known as meiotic arrest genes. These genes identify two independent pathways required for meiotic progression and spermatid differentiation. Gene expression analyses allowed definition of the most function, suggesting that it regulates the translation of twine (cdc25) mRNA, without affecting the spermatid differentiation pathway. My results also suggest that the centrosome phenotype of most mutants is likely to be an indirect consequence of meiotic arrest. Cytological analysis showed that premeiotic spermatocytes of fract mutants display 2 centrioles at each cell pole. However, meiotic ana-telophase I figures of the same mutants often exhibit two regular centrioles at one cell pole but only one at the opposite pole, suggesting that fract is required for centriole stability during the meiotic division. fract encodes a 322 aa testis-specific protein that contains WD repeats. The EMS-induced fract1 mutant allele I characterized carries a stop codon that truncates the Fract protein into a 298 aa polypeptide. A polyclonal antibody raised against Fract decorates the distal end of male meiotic centrioles. This specific staining pattern is lost in fract mutants, where the antibody decorates the entire centriole, suggesting that the C-terminal region of Fract is crucial for its correct localization. Collectively, these results strongly suggest that fract does not play an essential role in centriole duplication or assembly but it is instead required for the maintenance of centriole integrity during male meiosis, a function never described in other organisms.