Centrosome Maturation and Duplication in the C. elegans Embryo

National Institute Of Diabetes And Digestive And Kidney Diseases

Investigators

Linked publications, trials & patents

Abstract

Over the past few years, we have identified and characterized a number of genes with novel roles in regulating centrosome size and centriole duplication. Most of these genes were identified in a screen for factors that genetically interact with the kinase ZYG-1, a conserved master regulator of centrosome duplication. Analysis of individual szy genes has led to the identification of several molecular pathways that control centriole duplication by controlling the expression levels of centriole assembly factors. During the past year we completed a project on how the chromodomain helicase DNA-binding protein CHD-1 controls centriole assembly. We have found that deletion of the C. elegans chd-1 homolog suppresses the centriole duplication defect of the hypomorphic zyg-1(it25) mutant. We have also used CRISPR-based gene editing to construct an independent chd-1 null allele and have found that it also suppresses zyg-1(it25). Consistent with it know role as a chromatin remodeler, mutation of the ATPase domain of the endogenous protein leads to suppression. Further we have found that loss of chd-1 affects expression of certain components of the centriole assembly pathway. Specifically, loss of CHD-1 is accompanied by an increase in the total levels of SPD-2 and SAS-6 proteins. Also, loss of CHD-1 results in elevation of ZYG-1 levels at the centrosome. Using quantitative RT PCT, we have found that spd-2 and sas-6 mRNA levels appear unaffected, indicating that CHD-1 indirectly controls the expression level of these proteins. Our most recent efforts have focused on a genetic interaction between chd-1 and another zyg-1 suppressor. The transcription factor DPL-1, encoded by the dpl-1/szy-10 gene, also negatively regulates centriole duplication; similar to loss of chd-1, loss of dpl-1 both suppresses a zyg-1 hypomorphic allele and results in elevated SAS-6 levels. We have found that depletion of DPL-1 by RNAi can enhance suppression of zyg-1 by the chd-1(null) allele suggesting that the two genes function in distinct pathways. Interestingly we find that combining the chd-1(null) allele and dpl-1(RNAi) produces a synthetic embryonic lethality. We have characterized this interaction further and have found that chd-1;dpl-1 double mutants exhibit a strong embryonic lethality marked by the appearance of multipolar spindles, suggesting a centriole overduplication defect. Importantly, our work has revealed that CHD-1 like DPL-1 down regulates expression of CDK-2, a cyclin-dependent kinase known to promote centrosome duplication in vertebrates. Our work also shows that CHD-1 and DPL-1 appear to function independently of one another to down regulate CDK-2 and that loss of CDK-2 leads to downregulation of SAS-6. Overall, our results reveal a novel mechanism that controls centriole number. CHD-1 and DPL-1 cooperate to down regulate expression of CDK-2. CDK-2 in turn promotes expression of SAS-6 and centriole assembly. In a related project, we are employing biochemical and biophysical approaches to characterize centriole duplication; specifically, we have been investigating how ZYG-1 regulates (and is regulated by) downstream components of the centriole duplication pathway. SAS-5 and SAS-6 are coiled-coil-domain-containing proteins that form the structural scaffold of the centriole and require ZYG-1 for their incorporation into centrioles. These proteins are known to form dimers and higher order oligomers that are important for their function. However, the potential role of ZYG-1 in regulating the oligomeric state and intermolecular interactions of these proteins has not been investigated. We have expressed full-length recombinant proteins in E. coli and purified them to near homogeneity. We have found that ZYG-1 and SAS-5 physically interact in vitro and that ZYG-1 is capable of phosphorylating SAS-5 in vitro. Further we have found that other centriole components (SAS-6 and SAS-4) can modulate ZYG-1 kinase activity. Using mass spectrometry, we find that ZYG-1 phosphorylates SAS-5 on a number of highly conserved serine and threonine residues. Of particular interest, four of these phosphorylated residues are in a region predicted to bind SAS-4, suggesting that ZYG-1 might regulate SAS-4-SAS-5 interactions. To address this, we used site directed mutagenesis to create a series of non-phosphorylatable and phosphomimetic versions of SAS-5 and tested their ability to interact with both SAS-4 and ZYG-1. Strikingly, we found that SAS-4 and ZYG-1 bind to the same region of SAS-5 and that phosphorylation biases binding toward one or the other partner. That is, those phosphorylation events that promote SAS-4 binding, inhibit ZYG-1 binding, and vice versa. This suggests that a central event in centriole assembly involves a hand off of SAS-5 from ZYG-1 to SAS-4. To confirm that this mechanism is important in vivo, we have used CRISPR-based gene editing to mutate the phosphorylated residues in the endogenous sas-5 gene. Surprisingly, some of the mutants exhibit a cold-sensitive embryonic lethality marked by the appearance of monopolar spindles; this suggests that the phosphorylation status of these residues are important for centriole assembly. We are continuing to characterize this mechanism with the goal of obtaining a better understanding of how these factors interact on a molecular level to build a nine-fold symmetric centriole. Most recently we have taken a complimentary approach by using CRISPR to mutate conserved serine and threonine residues in SAS-5 to alanine and have found that serine 10 is essential for SAS-5 function. Further, we find that three serine residues (S331, 338, and 340) at the C. terminus of SAS-5 are also essential. Using multidimensional confocal imaging we have found that conversion of serine 10 to either alanine (non-phosphorylated mutant) or glutamate (phosphomimetic mutant) results in a sas-5 loss-of-function phenotype, whereby centriole duplication fails at a high rate. That the putative phosphomimetic S10E also presents as a loss of function indicates that a change to glutamate at this position does not fully mimic a phosphorylated residue. Using quantitative immunoblotting, we have found that the S10A and the S10E forms of SAS-5 are expressed at normal levels indicating that blocking phosphorylation of this residue results in a nonfunctional SAS-5 protein rather than an unstable protein. To provide evidence that SAS-5 is phosphorylated on serine 10 in vivo, we have raised an antibody that specifically recognizes serine 10 phosphorylated SAS-5. By immunofluorescence staining, we find that phospho-SAS-5 localizes to centrioles in a cell cycle dependent manner. Interestingly, the cell-cycle profile of phospho-SAS-5 at centrioles differs considerably from bulk SAS-5 protein. We are currently attempting to demonstrate that the phospho-SAS-5 signal is ZYG-1 dependent. Finally, we have also analyzed a triple mutant (S331A, S338A, and S340A) and found that it causes the production of excess centrioles. Quantitative western blots indicate that this SAS-5-A mutant is overexpressed two-fold relative to wild-type SAS-5. Interestingly, both the centriole overduplication defect and the over-expression defect can be rescued by co-expressing a wild-type version of SAS-5. This indicates that phosphorylation of these residues down-regulates expression of SAS-5 and that this mechanism can operate in trans to control unphosphorylated SAS-5 proteins.