Developmental Regulation of Centrosome Duplication

O'Connell, Kevin

Abstract

Centrosomes are the primary microtubule-organizing centers (MTOCs) in most cells and consist of a pair of centrioles wrapped within a cloud of pericentriolar material (PCM). During mitosis, each centrosome establishes one pole of the bipolar spindle. In Caenorhabditis elegans, the kinase ZYG-1 is essential for the duplication of centrioles. Embryos lacking maternal ZYG-1 activity fail to duplicate the paternally contributed centriole pair, and are thus unable to form bipolar spindles following first division. In contrast, loss of paternal ZYG-1 activity results in duplication failure during male meiosis, and the production of sperm with a single centriole. These sperm can still fertilize eggs but the resulting embryos assemble a monopolar rather than bipolar spindle at first division. These results demonstrate that ZYG-1 is required for centriole duplication during both the mitotic divisions of the embryo and the meiotic division of spermatocytes. Although ZYG-1 and other components of the centriole assembly pathway are absolutely required for centriole duplication during mitosis and meiosis, some recent data indicate that these factors are regulated differently during the two modes of division. We have found that small truncations of the c-terminus of ZYG-1 block centriole duplication during mitosis but drive the over-duplication of centrioles during meiosis. The behavior of these truncated forms of ZYG-1 seems to reflect their ability to localize to centrioles; the mutant proteins can accumulate at the meiotic centrioles of spermatocytes but are unable to localize to the mitotic centrioles of embryos. Similarly, we have found that the temperature-sensitive sas-6(or1167) mutation appears to differentially affect meiotic and mitotic centriole duplication. At the restrictive temperature, the sas-6(or1167) mutant strongly blocks male meiotic centriole duplication leading to an invariant monopolar spindle defect during the first embryonic division. In contrast, maternally-controlled mitotic centriole duplication is only blocked 60 percent of the time. Together these observations suggest that different cell types might utilize different mechanisms for regulating centriole number. During the past two years, we have been characterizing the effects of the sas-6(or1167) mutation on centriole assembly and stability. SAS-6 is a coiled-coil domain protein and a core component of the centriole. It is thought that centriole biogenesis involves the self-assembly of SAS-6 molecules into a nine-fold symmetric scaffold around which the centriole is constructed. In this model, the mother centriole only serves to define the site of centriole synthesis but does not play an active role in formation of the daughter. The sas-6(or1167) mutation changes an aspartate residue in the N-terminal head region to valine (SAS-6-D9V). Since the globular head of SAS-6 mediates the oligomerization needed to form a centriole scaffold, the or1167 mutation might affect the packing of the protein within the scaffold. Importantly, we have found that D9V form of SAS-6 leads to either failure of daughter centriole assembly or formation of a replication-compromised daughter. Immunoblotting experiments reveal that the protein encoded by sas-6(or1167) is expressed at a lower level than wild-type SAS-6. This defect is most severe at restrictive temperature, suggesting that the temperature-sensitive nature of this allele is due to an unstable protein. Recently, we have found that many of the monopolar spindles formed in sas-6(or1167) embryos lack core centriole comonents such as SAS-4 and SAS-6, indicating that the structure of the centriole is severely compromised. Similarly, in sas-6(or1167) male germ lines, we find that centriole markers are present in most cells prior to meiosis but are gradually lost as centrioles age. In collaboration with Thomas Muller-Reichert of MTC-Dresden, we have found by electron microscopy that SAS-6(D9V) centrioles are indeed abnormal in structure, and in some instances, appear to lack the nine-fold symmetry characteristic of normal centrioles. Surprisingly, despite this severe morphological defect, SAS-6(D9V)-based centrioles organize centrosomes and spindle poles of normal size. Thus, these structural defects to not impair all centriole functions. Most importantly however, we find that while mutant centrioles can direct centriole assembly using wild-type SAS-6 protein, they cannot do so using SAS-6(D9V) protein. Conversely, wild-type centrioles can direct centriole assembly with either form of SAS-6. Together, these findings provide evidence that the mother centriole plays an instructive role in daughter centriole assembly and that the molecular mechanism might involve resident SAS-6 in the mother centriole influencing the assembly of cytoplasmic SAS-6. As part of this project, we also are isolating genetic suppressors of the sas-6(or1167) allele using the same approach that we successfully employed to identify regulators of zyg-1. So far we have identified 41 independent sas-6(or1167) strains that can grow for multiple generations at the restrictive temperature. Six of the suppressors exhibit dominance while 23 appear recessive. One of the dominant suppressor mutations was found to be intragenic and results in a second missense mutation in the head region of SAS-6. It is likely that most, if not all, of the remaining suppressors carry a mutation in a gene other than sas-6 (extragenic suppressors); these can be used to identify potentially important meiosis-specific regulators of centriole number. In conjunction with the NIDDK Genomics Core Facility, we have conducted whole genome sequencing of 31 of the strongest suppressors. We have identified one suppressor as an allele of zyg-1. The zyg-1(bs84) mutation is in the SAS-6-binding domain suggesting that ZYG-1 might influence the packing of SAS-6 molecules in the centriole cylinder. We are currently exploring this possibility. This past year we have also completed a study of the function of the cyclin-dependent kinase CDK-11 in the germ line. CDK-11 has established roles in transcription, microtubule nucleation, and apoptosis. Recently a published report demonstrated an essential role for human CDK-11 in centriole duplication in somatic cells (Franck et al . (2011). PLoS ONE, 6(1), e14600). Our initial objective was to determine if CDK-11 functioned in a similar capacity in C. elegans and if possible to further dissect the role of CDK-11 in the centriole assembly pathway. C. elegans possesses two cdk-11 genes (cdk-11.1 and cdk-11.2). Deletion of cdk-11.2 did not produce an observable phenotype while deletion of cdk-11.1 caused a significant reduction in brood size. After extensive analysis, we did not find that the cdk-11 genes are required for centriole duplication in C. elegans, suggesting that this role is not universally conserved. Nevertheless, we found that CDK-11.1 is required for sperm and oocyte production. Interestingly, while CDK-11.1 is expressed uniformly throughout the germ line, CDK-11.2 is concentrated in germ line nuclei actively engaged in RAS-ERK signaling. The RAS-ERK pathway promotes the production and maturation of oocytes, and accordingly, we found that loss of cdk-11.1 led to a strong deactivation of this pathway. A similar but less severe effect was observed in the cdk-11.2 mutant. Thus, CDK-11.1 and CDK-11.2 cooperate to regulate oogenesis via MAP kinase signaling. While cdk-11.1 animals still produce some morphologically-normal gametes, they remain almost completely infertile. We therefore examined these mutant gametes and found evidence that sperm and oocytes fail to engage in cell-cell signaling during fertilization. Thus, CDK-11.1 is likely required for multiple aspects of gamete development.