Polycomb-group proteins are a well-characterized family of proteins involved in chromatin remodeling. In mammals, Polycomb-group proteins form two multiprotein complexes: Polycomb repressive complex 2 (PRC2), which represses gene expression, and PRC1.
PRC1, unlike its highly conserved sister, is a masterpiece of variety. It comprises the subunits Polycomb (Pc), Posterior sex combs (Psc), Polyhomeotic (Ph) and Sex combs extra (Sce) – but in humans there are five orthologs of Pc, six of Psc, three of Ph and two of Sce, which can, in theory, give rise to up to 180 different versions of PRC1. And while we don’t know for sure how many versions of PRC1 are to be found in human cells, one thing is clear: there doesn’t seem to be much difference in function between the orthologous complexes.
Why so many of them then? The reigning (but speculative) theory is that different versions of PRC1 may operate in different cell lineages and target different – cell-specific – sets of genes. There is some evidence to support this model: a number of the orthologs are predominant in particular cell types. But what we really lack is solid proof.
This week, Genome Biology publishes a study by Prof Gordon Peters and colleagues, who try to address this issue. Peters and colleagues perform a genome-wide study of PCR1 subunit localization in two different fibroblast lines, from which they identify co-localization between orthologs. And what they find is interesting: three Pc orthologs (CBX6, 7 and 8), and both Sce orthologs (RING1 and RING2), have surprisingly similar binding patterns to one another across the whole genome. So much for target specificity.
The study also looks at the binding patterns of 11 of the PRC1 subunit orthologs at representative loci and discover that these patterns, for all intents and purposes, are identical between orthologs, and often also between the two cell lines. And perhaps the most surprising discovery of all: there is no significant correlation between the localization of PRC1 subunits and gene expression of nearby genes.
Peters and colleagues conclude that different variants of PRC1 are not very specific after all. They speculate that different PRC1 variants co-operate to regulate gene expression, but that this regulation is subtle. And while solid proof of this new hypothesis is also lacking, the notion itself is a fascinating possibility.