In some cells in the eye, a GFP reporter gene near heterochromatin is expressed (green cells, right), but the adjacent white gene , which produces red pigmentation, is silenced (white cells, left), implying that gene promoters might respond individually to the repressive effects of heterochromatin. Credit: Reprinted from Cell 104, 839 – 847 © (2001), with permission from Elsevier Science

The transcriptional activity of a gene can be affected by its location in the genome. This phenomenon has been noted in many transgenic experiments and is also relevant to human disease, for example when the genomic location of a gene is altered by a chromosomal rearrangement. Position effects that reduce transcriptional activity can be mediated by heterochromatin — a compact and transcriptionally inert form of chromatin — which can silence genes placed within or near it. But position effects are not all-or-nothing phenomena, as genes seem to vary in their susceptibility to silencing by heterochromatin. Ahmad and Henikoff have now found a possible explanation for this variation, by showing that increased levels of a transcriptional activator for a gene that is silenced by heterochromatin can counteract the silencing effect.

Position effects were first discovered by Drosophila geneticists, who observed that mutations at certain positions in the fly genome give rise to mosaic phenotypes, the precise patterns of which vary between flies. To describe this phenomenon, they coined the term 'position effect variegation' (PEV). PEV reflects the incomplete silencing effect of heterochromatin and provides an opportunity to investigate factors that influence whether a gene is 'on' or 'off'.

Ahmad and Henikoff studied PEV in flies using a transgenic construct comprising two genes — a white gene that allows PEV to be easily visualized by variation in eye pigmentation, and a gene that expresses green fluorescent protein (GFP) under the control of the Gal4 transcriptional activator. The construct was located near a heterochromatic region and showed typical PEV. The authors then investigated whether altering the level of the Gal4 activator could affect the silencing of the construct's genes.

The results showed a clear effect not only on the GFP gene but also on the adjacent white gene — when Gal4 levels were increased, the degree of heterochromatic gene silencing was reduced. This led the authors to propose that PEV reflects a bistable equilibrium — a balance between the compact heterochromatic state and a more open conformation that is transcriptionally competent. The activator seems to promote the switch from the silent to the open state, which can extend beyond the gene that contains the binding sites for the activator.

But the conclusions go further. Although the effect of the activator can spread to the adjacent white gene, this is not always the case. In some flies, the white and GFP genes — which are only 2 kb apart — can be in different transcriptional states (see picture). This leads to the intriguing idea that heterochromatin might have an unsuspected fine structure, and might not be a block of uniform inactivity as generally supposed.