Abstract. To test the popular but unproven assumption that the metaphase-anaphase transition in vertebrate somatic cells is subject to a checkpoint that monitors chromosome (i.e., kinetochore) attachment to the spindle, we filmed mitosis in 126 PtK t cells. We found that the time from nuclear envelope breakdown to anaphase onset is linearly related (r 2 = 0.85) to the duration the cell has unattached kinetochores, and that even a single unattached kinetochore delays anaphase onset. We also found that anaphase is initiated at a relatively constant 23-min average interval after the last kinetochore attaches, regardless of how long the cell possessed unattached kinetochores. From these results we conclude that vertebrate somatic cells possess a metaphase-anaphase checkpoint control that monitors sister kinetochore attachment to the spindle.We also found that some cells treated with 0.3-0.75 nM Taxol, after the last kinetochore attached to the spindle, entered anaphase and completed normal poleward chromosome motion (anaphase A) up to 3 h after the treatment-well beyond the 9-48-min range exhibited by untreated cells. The fact that spindle bipolarity and the metaphase alignment of kinetochores are maintained in these cells, and that the chromosomes move poleward during anaphase, suggests that the checkpoint monitors more than just the attachment of microtubules at sister kinetochores or the metaphase alignment of chromosomes. Our data are most consistent with the hypothesis that the checkpoint monitors an increase in tension between kinetochores and their associated microtubules as biorientation occurs.T HE transition from metaphase to anaphase is a key cell cycle event that commits the cell to exit mitosis and enter a new interphase (reviewed in Murray, 1992;Sluder and Rieder, 1993). Since the equal segregation of chromosomes at mitosis is predicated on each acquiring a bipolar attachment to the spindle before anaphase onset, the metaphase-anaphase transition must be tightly coordinated with proper completion of chromosome attachment. Ensuring the essential coordination of these events would be relatively straightforward if the mitotic portion of the cell cycle was fixed in duration, but sufficiently long so that all chromosomes had time to acquire a proper attachment before disjoining. However, the stochastic nature of spindle formation makes chromosome attachment an unpredictable and errorprone process. In vertebrate cells a chromosome first attaches when one of its kinetochores, usually the one closest to and facing a spindle pole at nuclear envelope breakdown

Cultured primary human cells, which lack telomerase, enter a state of replicative senescence after a characteristic number of population doublings. During this process telomeres shorten to a critical length of approximately 5-7 kb. The mechanistic relationship between advanced cell passage, cellular senescence and telomeric function has yet to be fully elucidated. In the study described here, we investigated the relationship between changes in telomeric replication timing and/or sister chromatid separation at telomeric regions and advanced cell passage. Using fluorescence in situ hybridization, we analyzed the appearance of double hybridization signals (doublets), which indicate that the region of interest has replicated and the replicated products have separated sufficiently to be resolved as two distinct signals. The results showed that the replication and separation of several telomeric regions occurs during the second half of S-phase and that a delay in replication and/or separation of sister chromatids at these regions occurs in pre-senescent human fibroblasts. Surprisingly, in a significant percentage of pre-senescent cells, several telomeric regions did not hybridize as doublets even in metaphase chromosomes. This delay was not associated with extensive changes in methylation levels at subtelomeric regions and was circumvented in human fibroblasts expressing ectopic telomerase. We propose that incomplete replication and/or separation of telomeric regions in metaphase may be associated with proliferative arrest of senescent cells. This cell growth arrest may result from the activation of a mitotic checkpoint, or from chromosomal instability consequent to progression in the cell cycle despite failure to replicate and/or separate these regions completely.