tpohl

While it has been long known that centromeres of all Saccharomyces cerevisiae chromosomes reside in early replicating domains, increasing evidence indicates that this property is conserved among centromeres of fungi and some higher eukaryotes.  Surprisingly, little is known about the biological significance or the mechanism of early centromere replication; however, the extensive conservation suggests that it is important for chromosome maintenance.  We entertained two simple hypotheses: (1) centromeres ensure their early replication by promoting early activation of nearby origins; and (2) centromeres have migrated over evolutionary time to reside in early replicating regions.  Either hypothesis assumes that there is a selective advantage that early replication of centromeres provides the cell.  To distinguish between these two possibilities a variation of the Meselson/Stahl density transfer assay was employed to determine replication times of centromeres and their flanking origins in haploid strains harboring centromeres residing in either their endogenous position or translocated to a late replicating genomic region.  We have shown that a functional centromeres act in cis over a distance as great as 19 kb to advance the initiation time of origins, and that this effect depends on the centromere’s ability to establish a functional kinetochore (Pohl et al. 2012 PLoS Genetics). 

These observations strongly suggest that it is important for cells to maintain early centromere replication.  In light of these data, we hypothesize that centromeres need to be replicated early in S-phase to allow sufficient time for the assembly of functional kinetochores.  Therefore, delayed replication of centromeres might lead to increased chromosome loss.  To test this hypothesis, we set out to create a strain that has been modified to have a late replicating centromere and utilize this strain to answer the following questions: (1) can a centromere be made to replicate late? (2) is a chromosome harboring a late replicating centromere more unstable than its wild type counterpart?  (3) are cells harboring late replicating centromeres less fit than wild type cells?  (4) do late replicating centromeres evolve to be come early over evolutionary time?  Because centromeres regulate the activation time of adjacent origins, the simplest method for making a centromere late replicating is to knock out those origins.  To do so we have replaced ARS510 and ARS511 with different selectable markers in haploid cells.  The density transfer assay was used to show that centromere 5 on this modified chromosome is indeed late replicating.  Preliminary experiments subjecting diploids that are heterozygous for centromere 5 replication time indicate that having a late replicating centromere leads to increased chromosome rearrangements.