The retinoblastoma tumour-suppressor protein (pRb) and the E2F transcription-factor family function in cell growth and differentiation and do so by coordinating gene transcription. pRb can repress gene transcription by recruiting repressor complexes to E2F-bound gene promoters. Alexander Brehm, Nick Dyson and colleagues have now isolated from Drosophila melanogaster embryo extracts two such repressor complexes — which they named dREAM (Drosophila RBF, E2F and Myb-interacting proteins). Their results are reported in Cell.

D. melanogaster contains two pRb-family proteins, RBF1 and RBF2, and two E2F proteins, dE2F1 and dE2F2, which each form heterodimers with the transcription factor dDP. Chromatographic separation of D. melanogaster embryo nuclear extracts revealed three fractions that contained RBF1, of which one contained two separate complexes. Further purification revealed that these two complexes contained components of a dMyb complex that controls chorion gene amplification in follicle cells, as well as dE2F2, dDP and either RBF1 or RBF2.

Staining of polytene chromosomes with antibodies against various dREAM subunits showed extensive colocalization between these subunits. However, using an antibody against phosphorylated RNA polymerase II showed there was no overlap between the two staining patterns, which indicates that dREAM complexes associate with transcriptionally silent regions.

So, how does dREAM repress transcription? As dREAM complexes seem to lack chromatin-modifying enzymes, Brehm, Dyson and colleagues wanted to know whether dREAM complexes repress transcription by binding to histones directly. Indeed, several subunits bound the non-acetylated histone-H4 tail, but they failed to bind when the histone-H4 tail was acetylated. Given that transcriptionally silent chromatin is typically deacetylated, the authors propose that by binding to non-acetylated nucleosomes, dREAM complexes can protect them from modification, thereby maintaining the repressive state.

To identify the genes that are regulated by dREAM, the authors depleted dREAM subunits by treating D. melanogaster S2 cells with RNAi. Focusing on two sets of dE2F-regulated genes — cell-cycle-regulated A-group and E-group genes that have sex- and differentiation-specific expression patterns — they found that depletion of dREAM subunits de-repressed E-group genes, but not A-group genes.

Intriguingly, seven out of the eight components of dREAM were found to be related to the Caenorhabditis elegans synMuv class-B gene products, which are important for the development of the worm's male and female reproductive systems. Brehm, Dyson and colleagues propose that synMuv class-B proteins form a complex that, like dREAM, represses sex-related gene targets and so controls the worm's sexual development. So, the evolutionary conservation of pRb-specific repressor complexes is extensive and the authors predict that 'dREAM-like' complexes might also exist in mammals.