The spindle checkpoint ensures that newly born cells receive one copy of each chromosome by preventing chromosomes from segregating until they are all correctly attached to the spindle. The checkpoint monitors tension to distinguish between correctly aligned chromosomes and those with both sisters attached to the same spindle pole. Tension arises when sister kinetochores attach to and are pulled toward opposite poles, stretching the chromatin around centromeres and elongating kinetochores. We distinguished between two hypotheses for where the checkpoint monitors tension: between the kinetochores, by detecting alterations in the distance between them, or by responding to changes in the structure of the kinetochore itself. To distinguish these models, we inhibited chromatin stretch by tethering sister chromatids together by binding a tetrameric form of the Lac repressor to arrays of the Lac operator located on either side of a centromere. Inhibiting chromatin stretch did not activate the spindle checkpoint; these cells entered anaphase at the same time as control cells that express a dimeric version of the Lac repressor, which cannot cross link chromatids, and cells whose checkpoint has been inactivated. There is no dominant checkpoint inhibition when sister kinetochores are held together: cells expressing the tetrameric Lac repressor still arrest in response to microtubule-depolymerizing drugs. Tethering chromatids together does not disrupt kinetochore function; chromosomes are successfully segregated to opposite poles of the spindle. Our results indicate that the spindle checkpoint does not monitor inter-kinetochore separation, thus supporting the hypothesis that tension is measured within the kinetochore.
The spindle checkpoint monitors tension on chromosomes to distinguish between chromosomes that are correctly and incorrectly attached to the spindle. Tension is generated across a correctly attached chromosome as microtubules from opposite poles attach to and pull kinetochores apart, but are resisted by the cohesin that holds sister chromatids together. This tension generates separation between kinetochores as pericentric chromatin stretches and it also elongates the kinetochores. To monitor tension, the checkpoint could measure the separation between kinetochores or the stretch within them. We inhibited the ability of pericentric chromatin to stretch by tethering sister centromeres to each other, and we asked whether the resulting reduction in inter-kinetochore separation artificially activated the spindle checkpoint. Inhibiting inter-kinetochore separation does not delay anaphase, and the timing of mitosis was the same in cells with or without the spindle checkpoint, showing that the checkpoint is not activated. Inhibiting chromatin stretch does not alter the function of kinetochores as chromosomes are still segregated correctly, nor does it hinder the checkpoint. Cells whose sister kinetochores are held together can still activate the checkpoint in response to microtubule depolymerization. Our results indicate the spindle checkpoint does not monitor inter-kinetochore separation and likely monitors tension within kinetochores.