In a recent issue of PLoS Biology
, Marta Radman-Livaja, a post-doctoral associate in the Rando Lab
, used a novel histone-tagging technique to follow histone inheritance through multiple generations. She found that older histones tended to cluster near the 5′ end of genes (the “upstream” end that is closest to the promoter). To account for this pattern, the authors proposed a model in which histone dynamics are governed by three factors. While a cell is growing, histones turn over—that is, leave and rejoin nucleosomes. During gene transcription, surprisingly, histones are displaced from 3′ to 5′, a process the authors termed “passback.” During replication, histones spread in both directions, due to the mechanics of the replication machinery. In their model, and supported by mutation experiments that disrupted these various effects, the net accumulation of older histones at the 5′ end is largely due to passback.
This dynamic model suggests that histone-based epigenetic information cannot be inherited in a pinpoint fashion, since individual histones or whole nucleosomes are not precisely relocalized to their original locations, either after replication or after transcription. Instead, as other experiments have suggested, the minimal “unit” of epigenetic inheritance is likely to span multiple nucleosomes, enough to cancel out the noise from these three conflicting factors. Future experiments will address whether the two replicated DNA strands handle histone movements differently, and whether mutations can be identified that influence histone movements during replication.