Chromatin insulators are defined as transcriptionally neutral elements that prevent negative or positive influence from extending across chromatin to a promoter. Here we show that yeast subtelomeric anti-silencing regions behave as boundaries to telomere-driven silencing and also allow discontinuous propagation of silent chromatin. These two facets of insulator activity, boundary and silencing discontinuity, can be recapitulated by tethering various transcription activation domains to tandem sites on DNA. Importantly, we show that these insulator activities do not involve direct transcriptional activation of the reporter promoter. These findings predict that certain promoters behave as insulators and partition genomes in functionally independent domains.

DNA repetitions may provoke heterochromatinization. We explore here a model in which multiple cis-acting sequences that display no silencing activity on their own (protosilencers) may cooperate to establish and maintain a heterochromatin domain efficiently. Protosilencers, first defined in budding yeast, have now been found in a wide range of genomes where they appear to stabilize and to extend the propagation of heterochromatin domains. Strikingly, isolated or moderately repeated protosilencers can also be found in promoters where they participate in transcriptional activation and have insulation functions. This suggests that the proper juxtaposition of a threshold number of protosilencers converts them from neutral or transactivating elements into ones that nucleate heterochromatin. Interactions might be transient or permanent, and are likely to occur over distances by looping. This model provides a conceptual framework for as varied phenomena as telomere-driven silencing in Drosophila, X inactivation in mammals, and rDNA silencing in S. cerevisiae. It may also account for the silencing that occurs when multiple copies of a transgene are inserted in tandem.

We have designed a modified version of the Dam identification technique and used it to probe higher-order chromatin structure in Saccharomyces cerevisiae. We fused the bacterial DNA methyltransferase Dam to the DNA-binding domain of TetR and targeted the resulting chimera to Tet operators inserted in the yeast genome at the repressed locus HML. We then monitored the methylation status of HML and other sequences by a quantitative technique combining methylation-sensitive restriction and real-time PCR. As expected, we found that TetR-Dam efficiently methylated HML in cis. More strikingly, when TetR-Dam was present at HML, we observed increased methylation in the III-L subtelomeric region but not in intervening sequences. This effect was lost when the HML silencers were inactivated by mutations. When the HM silencers and the Tet operators were transferred to a plasmid, strong methylation was clearly observed not only in the III-L subtelomeric region but also at other telomeres. These data indicate that HM silencers can specifically associate with telomeres, even those located on different chromosomes.