For a long time, heterochromatin was considered to be the 'dark matter' of the genome: highly condensed, late replicating and associated with lack of gene expression. Various lines of evidence indicated that heterochromatin could lead to transcriptional silencing, but the mechanism was unknown. Studies of position-effect variegation (PEV) in fruitflies, and of the centromeric and mating-type regions in fission yeast, have proved key to shedding light on the mysteries of heterochromatin structure and function.

PEV — a clonally inherited pattern of active and silent gene states — was first described by Hermann Muller in 1930. Later, it was shown that affected silenced genes had been translocated to lie in close proximity to heterochromatin. In 1990, Sarah Elgin and colleagues identified heterochromatin protein 1 (HP1) and showed that mutations in the gene that encodes this protein in fruitflies behave as dominant suppressors of PEV, implicating HP1 in generating normal chromatin structure.

Five years later, the same group showed that altered chromatin packaging had a role in PEV. They used a mobile transposable element — a popular tool among fruitfly geneticists — to insert a marker gene at various locations throughout the genome. PEV was seen for transgenes that inserted near centromeres, telomeres and on chromosome IV, which is heterochromatic in the fruitfly. Careful analysis of transgene expression and chromatin structure showed that loss of expression was associated with an altered nucleosome array and reduced accessibility of restriction enzymes to the promoter. So, altered chromatin conformation underlies PEV, at least around the centromeres.

Studies of pericentromeric chromatin in fission yeast brought some important revelations about heterochromatin formation seven years later. The groups of Robert Martienssen and Shiv Grewal found that deleting genes that encode key components of the RNA interference (RNAi) machinery resulted in accumulation of complementary transcripts from centromeric heterochromatic repeats. Expression of centromeric transgenes in these deletion strains was de-repressed and histone H3K9 methylation — a mark of heterochromatin — was lost. These results raised the possibility that the RNAi machinery targets specific histone modifications to heterochromatic regions.

In a publication that followed two weeks later, the Grewal group turned to the mating-type region of fission yeast to study the factors that favour heterochromatin formation in cis. Little was known about them, except that repeat elements tended to be heterochromatic. This was an important issue — as was known from PEV, the heterochromatic state could spread to inactivate nearby genes. The authors found that cenH, a centromere-homologous repeat that normally lies at the silent mating locus, is sufficient to bring about heterochromatin formation at ectopic sites. This ability is associated with H3K9 methylation and recruitment of Swi6 (a yeast HP1 counterpart) and, importantly, requires the RNAi machinery.

The dark matter of the genome was gradually becoming illuminated. The encouraging news was that the findings in fruitflies and fission yeast fitted with results from mammalian cells, in which we now know that the targeting of epigenetic modifications to repeats by the RNAi machinery also has a central role in heterochromatin formation and gene silencing.