Determining the mechanism of the CSR-1 RNAI pathway in organizing holocentromeres in Caenorhabditis elegans

Abstract

During cell division, DNA has to be equally separated to the daughter cells. Errors during this process can lead to unbalanced genetic material which is a common cause of genetic disorders and a hallmark of cancer. The chromosomal region that is captured by the microtubules to pull the duplicated sister chromatids apart is called the centromere. Most functional centromeres are marked by a conserved histone H3 variant CENP-A, but are not defined by specific DNA sequences. Centromeres can occasionally form at ectopic, non-centromeric DNA loci leading to genome instability. Therefore, proper CENP-A organization is functionally important, yet the underlying mechanism remains unclear. In holocentric nematode C. elegans, CSR-1 RNA interference (RNAi) pathway is required for proper chromosome segregation. The argonaute CSR-1 associates with 22G-short interference RNAs (siRNAs) to target endogenous germline transcripts, to protect them from being silenced by other RNAi pathways and to fine-tune targets’ gene expression. These germline gene loci are anti-correlated to the DNA sequences where worm centromere protein HCP-3 (CENP-A) occupies. However, the causational relationship between CSR-1 RNAi pathway and the chromatin distribution of HCP-3 remains unclear. By live imaging, CSR-1-depleted embryos exhibited a delay in mitotic exit, a reduction in spindle pole separation, a reduction in spindle pole repulsion on mitotic chromosomes and chromatin HCP-3 mislocalization. These phenotypes suggested that the chromosomes in the CSR-1-depleted embryos are likely misattached by the microtubules, which could lead to chromosome missegregation. In particular, CSR-1 was found to regulate HCP-3 chromatin localization. In the CSR-1-depleted embryos, HCP-3 was neither over-expressed nor over-stabilized on the chromatin. However, these embryos had more HCP-3 on the poleward-faces of the chromosomes which corresponds to denser HCP-3 foci that appeared on stretched chromatin fibers. Moreover, CSR-1 might suppress HCP-3 loading to ectopic chromatin region. HCP-3 was found to localize onto chromatin in CSR-1-depleted embryos independent of the HCP-3 licensing factor KNL-2. Such ectopic HCP-3 may be functional to assemble more kinetochore proteins, as HCP-4 and NDC-80 localizations at the kinetochore were also increased. CSR-1 was validated to regulates HCP-3 through its involved RNAi pathway. The RNaseH-related domain of CSR-1 was shown to be important for the HCP-3 restriction, whereas the N-terminal extension of the longer CSR-1 isoform (CSR-1a), the endogenous mRNA sequence of hcp-3, and the CSR-1-antagonized HRDE-1 RNAi pathway were not required. CSR-1 may suppress HCP-3 occupancy by either targeting an unidentified negative regulator of HCP-3 or by depositing an inheritable epigenetic mark at its targets. To determine whether CSR-1’s proximity to germline-expressed chromatin regions through its binding to nascent mRNA affects HCP-3 localization, it will be important to check the HCP-3 occupancy at the locus where CSR-1 is artificially tethered. The results suggest that CSR-1, together with its RNAi pathway prevents ectopic HCP-3 chromatin localization. It highlights a novel role of CSR-1 in epigenetic regulation of HCP-3 arrangement to orchestrate chromosome segregation. The study has promoted our understanding of the spatial specification of centromere protein and the centromere identity in C. elegans and potentially in other holocentric organisms.