When a cell replicates its genetic material, divvies up its chromosomes, and then separates into two new, daughter cells, a lot can go wrong. Mistakes in this process can be biologically costly—leading to genetic mutations, cell death, or cancer—so living cells have evolved ways to ensure that the steps of mitosis proceed in the right order and at the right speed. Molecular checkpoints detect the state of a cell by measuring levels of signaling molecules, tension, or DNA damage, and only send signals for mitosis to proceed when certain conditions are met. Now, in three separate papers simultaneously published in Current Biology, scientists have examined the details behind one checkpoint, and discovered that the inactivation of a checkpoint may be as important—and complex—as activation.
Early in mitosis, chromosomes are pinned onto microtubules and centered in a mitotic cell, with the arms of the mitotic spindle stretched to each end. “[Tension] is probably the most important mechanism to ensure that daughter cells inherit the right number of chromosomes,” said Mark Petronczki, a cell biologist at Cancer Research UK and senior author of one of the new papers. If tension on the kinetochore—where chromosomes attach to the spindle assembly—is lost during metaphase, a spindle attachment checkpoint managed by the anaphase-promoting complex/cyclosome (APC/C) keeps mitosis from proceeding.
But once mitosis has successfully moved from metaphase to anaphase, sister chromatids must separate, and tension will be lost. Petronczki wanted to know why the spindle attachment checkpoint didn’t stop mitosis in anaphase when this separation occurred.
Petronczki knew that APC/C degraded two proteins to allow mitosis to move from metaphase to anaphase—securin, which is responsible for the actual physical attachment between sister chromatids, and cyclin B, which pairs with the kinase Cdk1. He and his colleagues wondered whether cyclin B’s degradation, or the resulting inactivation of Cdk1, played a role in turning off the spindle assembly checkpoint as a cell moved to anaphase. So they created a version of cyclin B that couldn’t be degraded by APC/C and compared it to the wild type protein in HeLa cells.
Indeed, in the cells with non-degradable cyclin B, sister chromatids separated, but then mitosis stalled, in a new state that Petronczki’s group dubbed “pseudoanaphase.” Without the degradation of cyclin B, as cells moved into anaphase, it seemed that the spindle checkpoint remained active.
To test whether Cdk1—rather than cyclin B on its own—was carrying out the task of mediating the checkpoint, Petronczki next applied a Cdk1 inhibitor to cells and discovered that the checkpoint didn’t engage (1).
“Scientists generally think about the checkpoint as in one of two states: either engaged or satisfied,” said Petronczki, “but what our paper shows is that there are times when it is in a different inactivated state.”
Too Slow to Make a Difference?
While Petronczki’s group was carrying out their experiments on HeLa cells, Silke Hauf, a biologist at the Tuebingen Max Planck Institute, now at Virginia Tech, was studying the same checkpoint in yeast. “What I really wanted to look at was the kinetics of checkpoint signaling,” said Hauf. “It has always been assumed that cells need to very quickly realize whether there’s a problem or not.” To test just how quick this reaction is, Hauf focused on the spindle assembly checkpoint.
Like Petronczki, Hauf began with non-degradable cyclin B, but she paid particular attention to the timing of the events that followed. While checkpoint proteins began to accumulate at the kinetochores, and mitosis seemed to stall, APC/C was never inhibited—the true mark of the checkpoint being activated. Hauf wondered whether mitosis was just progressing so slowly that it never reached the stage of inhibiting APC/C. So her team used genetic tricks to change the speed the cell was progressing through different steps of mitosis.
When the authors slowed down the detachment of sister chromatids, cells had more time to turn on the checkpoint, and it was eventually activated. But it took up to 16 minutes; in normal anaphase, sister chromatids have already fully separated after 5 or so minutes (2). This means that, in normally progressing cells, there’s never time for the checkpoint to turn on.
While Petronzcki’s group concluded that no checkpoint activation during anaphase meant the checkpoint was turned off at the metaphase to anaphase transition, Hauf doesn't quite agree.
“There’s a lot of data that the checkpoint is shut off in anaphase,” said Hauf. “But I think the question that still isn’t answered is when exactly is it shut off? And what our data says is that it’s not being shut off immediately.” Instead, it might be progressing at a very slow rate.
“I don’t think the discrepancies between our papers are in the data,” Hauf said. “I think they’re more in the way that we each like to interpret data.” Petronczki suggested that the timing of checkpoint activation, deactivation, and responses could vary between organisms.
More Supporting Evidence
For the past few years, biochemist Béla Novák of Oxford University and colleague Kim Nasmyth have been studying the spindle assembly complex in mice. When they heard that Petronczki’s group was getting ready to submit a paper on the same topic, they decided to wrap up their findings and submit a paper alongside his.
Rather than beginning with cyclin B, Novák’s group wanted to confirm that APC/C was responsible for the spindle assembly checkpoint in the first place. They knew that two proteins—Cdc20 and Cdh1—activate APC/C during mitosis, so they genetically engineered a line of mice that lack Cdc20. When they analyzed oocytes attempting to copy genetic material from these mice, they found that without Cdc20 to activate APC/C, cells couldn’t move to anaphase and separate sister chromatids.
“An obvious solution is that APC/C normally destroys something which is critical in prometaphase but absent in anaphase,” said Novák.
When the researchers added a Cdk1 inhibitor to the eggs, sister chromatids could separate. They concluded that Cdk1 was the key substrate that needed to be degraded to move cells into anaphase and allow the release of tension on the kinetochores.
Through a series of further experiments, Novák and his colleagues confirmed that Cdk1 and cyclin B were critical to turning off the spindle assembly checkpoint in an irreversible, switch-like way (3).
Together, the three new papers build a strong case for the importance of the Cdk1 and cyclin B complex in moving cells to anaphase and keeping the checkpoint from activating when sister chromatids separate. But the scientists all agree that questions about the transition remain. How is microtubule stability regulated? How are all cell cycle checkpoints set up at the launch of mitosis and destroyed at the end? And what is the role of kinetics in each species?
“This is how cells avoid getting into terrible messes like cancer,” said Petronczki, “and we still don’t know a lot about it.”
1. Vazquez-Novelle, M.D., Sansregret, L., Dick, A.E., et al. (2014) Cdk1 Inactivation Terminates Mitotic Checkpoint Surveillance and Stabilizes Kinetochore Attachments in Anaphase. Current Biology 24:638-545.
2. Kamenz, J., Hauf, S. (2014) Slow Checkpoint Activation Kinetics as Safety Device in Anaphase. Current Biology 24:646-51
3. Rattani, A., Vinod, P.K., Godwin J., et al. (2014) Dependency of the Spindle Assembly Checkpoint on Cdk1 Renders the Anaphase Transition Irreversible. Current Biology 24:630-637