During mitosis, duplicated chromosomes are separated and segregated into two daughter cells. This is achieved by the action of spindle microtubules. In metaphase, microtubules radiating from centrosomes at two poles in the cell, attach to the condensed chromosomes at proteinaceous structures called kinetochores. The daughter chromosomes are then pulled slowly towards opposite poles during anaphase. Mostly, eukaryotes have monocentric chromosomes – meaning that each chromosome contains one centromere. Kinetochores are associated with domains of heterochromatin close to the centromeres. This pericentric heterochromatin is composed of repetitive sequences, and is generally transcriptionally silenced.
An important feature of the mechanisms stabilising the heterochromatic state in some organisms are small RNAs. In the fission yeast Schizosaccharomyces pombe an RdRP-dependent population of small RNAs, directed against repetitive sequence, has been shown to target pericentromeric heterochromatin, stabilising the centromeres. Perturbations to this pathway therefore lead to mitotic chromosome segregation defects.
Not all organisms’ chromosomes contain single centromeres. Nematodes have holocentric chromosomes in which multiple spindle attachments are made into continuous kinetochores spanning the length of the chromosomes. The assembly of centromeres and kinetochores in both monocentric and holocentric chromosomes have many conserved features. For instance, the histone variant HCP-3/CENP-A is found in centromeric nucleosomes. Unlike in monocentric chromosomes though, in nematodes it is incorporated into the whole poleward face of condensed chromosomes.
In C. elegans, a number of mutations affecting components of RNAi pathways were found to cause mitotic and meiotic chromosome segregation defects. As discussed in the previous post, these included drh-3 (encoding a helicase necessary for the biosynthesis of 22G-RNAs), and the Argonaute protein encoding csr-1. Claycomb et al. found that mutations in two additional genes, the RdRP encoding ego-1, and the tudor domain protein encoding ekl-1, caused similar defects. These included oocytes with abnormal complements of chromosomes, underproliferated germlines, and high incidence of males (him) phenotypes. Partial mutant alleles of csr-1 or drh-3 had low proportions of viable progeny, with many worms dying at various points in embryogenesis.
In embryos depleted of any of these factors, chromosomes initially condense normally in prophase. However, during metaphase the chromosomes fail to align into ‘plates’ perpendicular to the long axis of the spindles, and at anaphase the researchers observed chromosomal bridging across the midzone of the spindle. This led to the accumulation of cells with abnormal chromosomal complements, and the death of embryos.
Claycomb et al. then looked at the distribution of the centromeric histone variant HCP-3/CENP-A in RNAi depleted embryos. They found that although it was loaded onto chromosomes, instead of the normal poleward localisation, it was distributed in a disorganised pattern over the metaphase chromosomes. Similarly disorganised patterns were seen with an array of centromere and kinetochore proteins, as well as proteins necessary for chromosomal condensation and cohesion.
Deep sequencing of the small RNAs associated with CSR-1, showed that this argonaute binds 22G-RNAs targeted against protein-coding genes. This population of 22G-RNAs are dependent upon drh-3, ego-1, and ekl-1 and were antisense to at least 4191 genes. Interestingly, in glp-4 mutants that lack a germline, 22G-RNAs corresponding to ~80% of the CSR-1 target mRNAs were strongly depleted. This suggests that the CSR-1 associated 22G-RNAs originate in the germline and target genes expressed in the germline. No differences were found in the transcriptional profiles of csr-1 mutants and wild type worms, showing that CSR-1 22G-RNA system does not work by silencing its’ targets.
During the development of the germline, DRH-3, EGO-1 and CSR-1 all localise to P-granules – perinuclear nuage structures important for germline specification and small RNA mediated activities. In later stages of oocyte maturation CSR-1 becomes enriched in nuclei. In mitotic cells all four factors were enriched in prophase nuclei, and could be observed to localise to metaphase chromosomes. Using chromatin immunoprecipitation (ChIP), Claycomb et al. demonstrated that CSR-1 directly bound genomic loci targeted by CSR-1 associated 22G-RNAs.
These slightly disparate lines of evidence can be condensed into a basic model by which the CSR-1 22G-RNA pathway contributes to the regulation of chromosome segregation. CSR-1 associated 22G-RNAs are generated in the germline. Their biogenesis occurs in P-granules and depends on the RdRP complex (comprised of the helicase DRH-3, the RdRP EGO-1, and the tudor domain protein EKL-1) acting upon transcripts of germline expressed genes. Guided by these 22G-RNAs, CSR-1 translocates to the nucleus, and via chromatin modification defines chromosomal domains. It appears that there is an inverse relationship between domains targeted by CSR-1 22G-RNAs and those enriched for the centromeric histone variant HCP-3/CENP-A. Hence, CSR-1 marks domains adjacent to centromeric domains. As CSR-1 22G-RNA targets are distributed relatively uniformly across the genome, the defined domains can serve to position kinetochores along the length of the chromosomes. CSR-1 22G-RNAs are maternally inherited, and the CSR-1 defined chromatin domains are epigenetically stable, enabling the correct assembly of kinetochores during mitoses throughout development.
It is interesting to note that two related 22G-RNA pathways, those associated with WAGO or CSR-1 argonautes, with a similar biogenesis pathway, can have such diverse cellular roles. As we have seen the WAGO 22G-RNA pathway acts to silence deleterious transcription. The CSR-1 22G-RNA pathway doesn’t silence its’ targets but instead intricately organises chromosome structure to ensure proper segregation. Does this pathway have roles beyond this function? In later posts in this series I’ll cover papers that suggest it may also participate in a global genome surveillance network.
See also a post on a paper that gives a radically different take on the CSR-1 pathway: The CSR-1 siRNA pathway gets more enigmatic
Important ideas regarding the role of the CSR-1 pathway being involved in epigenetic licensing are discussed in these two posts:
Interacting small RNA pathways in worms 5: Global Genome Surveillance
Epigenetic Licensing of a Sex Determination Gene
Claycomb JM, Batista PJ, Pang KM, Gu W, Vasale JJ, van Wolfswinkel JC, Chaves DA, Shirayama M, Mitani S, Ketting RF, Conte D Jr, & Mello CC (2009). The Argonaute CSR-1 and its 22G-RNA cofactors are required for holocentric chromosome segregation. Cell, 139 (1), 123-34 PMID: 19804758