Credit: © Josefbosak |

The list of biological processes that are controlled by chromatin structure or small RNAs has expanded rapidly in recent years, and a few examples are emerging in which these two modes of regulation are connected. Research presented in three companion papers now links an RNAi pathway to chromosome structure and segregation.

The structure of chromatin close to centromeres is thought to influence the formation of kinetochores — protein structures that are essential for chromosome segregation. Most eukaryotes have a single centromere per chromosome, and pericentromeric repetitive sequences help to define their chromatin structure. Nematodes are unusual: their chromosomes are 'holocentric' — that is, they have centromeres and kinetochores along their length. So how are holocentric centromeres defined?

Previous studies have shown that in Caenorhabditis elegans, depletion of two RNAi pathway proteins — the Argonaute protein CSR-1 or the Dicer-related helicase DRH-3 — caused chromosome segregation defects. Claycomb and colleagues found that loss of two other RNAi factors, the RNA-dependent polymerase EGO-1 or the Tudor domain-containing protein EKL-1, had similar effects. They found that depletion of any of these four factors caused dramatic disorganization in the chromosomal localization of proteins that are vital for chromosome segregation, including condensin, cohesin, the centromeric histone variant HCP-3 and kinetochore factors.

Argonaute proteins interact with endogenous small interfering RNAs (siRNAs) and, by deep sequencing, the authors found that CSR-1 associates with a subset of siRNAs called 22G-RNAs. Gu et al. showed that DRH-3, EKL-1 and EGO-1 are required for 22G-RNA biogenesis and that the majority of these RNAs are antisense to protein-coding genes. Strikingly, however, depletion of CSR-1 or DRH-3 did not significantly alter the expression of the gene targets of CSR-1-associated 22G-RNAs, which suggests that these siRNAs might function primarily in chromosome segregation. An effect of CSR-1-associated 22G-RNAs and chromosome segregation, rather than gene expression, was also confirmed by van Wolfswinkel and colleagues, who found that the levels of 22G-RNAs are controlled by the nucleotidyltransferase CDE-1; the absence of CDE-1 causes 22G-RNA accumulation and chromosome segregation defects.

Claycomb and colleagues found that 22G-RNAs guide CSR-1 to the target genes, which are spaced evenly along nematode chromosomes in a pattern similar to the distribution of holocentric centromeres. The authors propose that CSR-1–RNA complexes might recruit chromatin modifiers that define centromeric domains and therefore control the position of kinetochores to ensure orderly segregation. This mechanism is analogous to the RNAi pathway that mediates pericentromeric heterochromatin formation in fission yeast, but different in that coding rather than repetitive sequences are targeted. The similarities between these processes in nematodes and yeast raise the exciting possibility that interlinked siRNA and chromatin regulatory pathways might be found throughout eukaryotic species.