After 40 years of searching for the eukaryotic replicator sequence, it is time to abandon the concept of 'the' replicator as a single genetic entity. Here I propose a 'relaxed replicon model' in which a positive initiator-replicator interaction is facilitated by a combination of several complex features of chromatin. An important question for the future is whether the positions of replication origins are simply a passive result of local chromatin structure or are actively localized to coordinate replication with other chromosomal activities.In the summer of 1962, François Jacob and Sydney Brenner sketched in the sand of France's La Tranche-sur-Mer a model (FIG. 1a) to explain the regulation of DNA synthesis in bacteria 1 . Their 'replicon model' proposed that replication was regulated by a positive interaction between an initiator protein, which is encoded by a structural gene, and a specific genetic element of recognition, which is known as the replicator. Although originally intended to explain the replication of circular bacterial chromosomes, the replicon model was quickly adopted as the paradigm for eukaryotic chromosomes as well (FIG. 1b). Proteins with the properties of initiators have been identified in species from Escherichia coli to man. By contrast, genetically defined replicators have yet to be identified in most eukaryotes and are clearly dispensable for replication in many eukaryotic systems. Even in its shortcomings, though, the model has been useful. Indeed, the replicon model was the driving force behind the discovery of bacterial and budding yeast replicators as well as initiator proteins -from dnaA in prokaryotes 2 to the origin-recognition complex (ORC) in eukaryotes 3 . Indeed, all cellular chromosomes can be thought of as replicons, or clusters of replicons -the replication of which is regulated, at least in part, by a positive initiator-replicator interaction.In this historical perspective, I will summarize the salient findings over the course of these four decades (TIMELINE), with a focus on strategies to identify metazoan replicators and the debate over their existence. I propose that we replace the concept of the replicator as a specific DNA sequence with that of a context-dependent element that is recognized by the initiator through several features, which can include DNA sequence. By embracing the complexity of replicators as being akin to that of promoter elements, future experiments can define these various features that influence the initiator-replicator interaction and address the biological significance of specifying replication-origin sites, given that site-specific initiation is dispensable for regulated genome duplication 30 .
The power of predictionOne of the beauties of the replicon model was the conceptually straightforward predictions that it provided. A replicator can confer the ability to replicate onto any DNA segments that are Competing interests statement