The team was not initially looking for a replication completion checkpoint. They were studying the function of a yeast complex called Smc5-Smc6 that had been suggested to promote DNA recombination and repair.
Budding yeast that lack Smc5-Smc6 do not survive more than a few cell cycles, so the team synchronized yeast cells in G1, knocked out the complex, and then observed the yeast over one cell cycle. S phase and mitosis appeared to be normal. In the subsequent interphase, however, the Rad53 DNA damage signal was activated. This damage response, the team discovered, was due to a failure of chromosomes to separate correctly at anaphase.
Thinking that this nondisjunction might be caused by a failure to resolve recombination events, the team knocked out critical recombination genes to see whether the problem was bypassed. Loss of recombination only slightly rescued the phenotype, however.
If Smc5-Smc6 was knocked out after S phase, but before metaphase, chromosomes segregated normally, showing that the complex is required during (or before) DNA replication. Loss of Smc5-Smc6 before S phase caused replication forks to persist into metaphase. Labeling of nascent DNA in these cells revealed that large regions of rDNA were still unreplicated. rDNA is a major binding site for Smc5-Smc6 and was also the main site of the nondisjunction.
Aragón suggests that Smc5-Smc6 might act as a chromatin structure modifier that allows the replication machinery to progress unhindered. Despite the slowed replication in the absence of Smc5-Smc6, the cells continued with cell division.
Yeast lack this replication completion checkpoint because they might not need it. There are many unused replication origins that can become active during late S phase, explains Aragón. Thus, it's likely that normal cells can easily replicate their DNA in time for metaphase.