Results: We first generated a high-resolution localization map of all to-date discovered proteins in the worm centriole using Ultrastructure Expansion coupled with STimulated Emission Depletion microscopy (U-Ex-STED, with an effective lateral resolution of ~14 nm), as well as electron microscopy (EM) and electron tomography (ET). All 12 components analysed localize to the centriole in a ring-like distribution with distinct diameters and often display 9-fold radial symmetry. By overlaying U-Ex-STED and EM images using the radial array of microtubules as a ruler, we could map each centriolar protein to a given ultrastructural compartment (Fig 2) .
We then tested how centriolar ultrastructure is altered over the course of oogenesis elimination using EM and U-Ex-STED. Additionally, in order to investigate compositional changes of individual centriolar proteins during elimination, we conduct two-color live imaging of endogenously tagged GFP or RFP-tagged centriolar proteins using a custom designed microfluidic device. Combining these approaches, we found that the evolutionarily conserved central tube protein SAS-1, which bundles microtubules when ectopically expressed in human cells 5, is the first component to be lost from centrioles, starting at the transition to late prophase. This is accompanied by central tube loss as revealed by EM. This is followed by progressive widening of the centriole and sequential loss of centriolar components, starting with microtubules and then proteins such as SAS-4 or SPD-2.
Since SAS-1 is the first protein to depart from the centriole during oogenesis elimination, we analyzed this process in sas-1 mutants. It was previously reported that centrioles form in sas-1 mutant sperm, but that they are dismantled in the zygote after fertilization. Importantly, we found that centriole widening and disintegration of centriolar components occurred prematurely during oogenesis in sas-1 mutants.
To gain insights into the machinery that executes centriole elimination, we inactivated candidates using RNAi, as well as null or conditional mutant alleles, and acute pharmacological inhibition using a recently published cuticular mutant gmap-1(ulb13), which renders the worm permeable to small molecules. This analysis revealed that inhibition of the proteasome leads to a significant delay in centriole elimination.
We then found that the 20S proteasome and ubiquitin were not detected at centrioles during early prophase. Strikingly, however, we found ubiquitin in the vicinity of centriolar microtubules and SAS-4 at the onset of late prophase, concomitant with centriole widening. Moreover, found the 20S proteasome transiently localizing to centrioles thereafter, at the time we observed loss of structural integrity of centriolar components. Importantly in sas-1 mutants, decoration of centrioles with ubiquitin and proteasome localization to centrioles occurred prematurely.
To test if centriole elimination mechanisms are widely conserved, we established Naegleria gruberi as a model system in the lab. This protist, which is distantly related to animals, switches between a mode in which it propels itself with two flagella, each templated by one centriole, and an amoeboid mode in which centrioles, flagella and MTs are absent. Using pharmacological inhibition, we found tha centriole elimination can be blocked by proteasome inhibition. Moreover ubiquitin and the 20S proteasome are becoming enriched over the course of elimination also in N. gruberi, potentially hinting to a conserved origin of the process of centriole elimination.
Our results could have important implications for human reproductive health. In a future work we will screen the pathways here uncovered to be involved in the process for their involvement in female fertility. Moreover, having established a novel way for pharmacological inhibition in C. elegans using gmap-1 mutant animals will allow to use the germline and the early embryo of WT and mutants to screen through small molecule libraries for the purpose of e.g. drug repurposing.
The results obtained are partially published or will be published in peer-review journals and pre-print server (bioRxvi) and were presented orally on several scientific meetings.
