Simply effective

The mechanisms that underlie biological processes often turn out to be more complex than was initially thought. However, a recent study of the spindle checkpoint, which was reported by Hongtao Yu and colleagues in Molecular Cell, hints at a regulatory mechanism of striking simplicity.

The spindle checkpoint ensures the accurate separation of chromosomes by blocking the ubiquitin-ligase activity of the anaphase-promoting complex/cyclosome (APC/C) in response to improper chromosome alignments and inappropriate tension of the microtubular spindle. Many proteins are involved in controlling this checkpoint, so its regulation is complex. In particular, the Bub1 kinase interacts with and regulates several other checkpoint proteins and was expected to be an upstream component of the checkpoint. But, Yu and colleagues now show that, surprisingly, Bub1 inhibits the activity of APC/C directly.

The authors first showed that a significant fraction of HeLa cells that had been depleted for Bub1 using RNA interference (RNAi) failed to undergo mitotic arrest after treatment with nocodazole (a spindle-damaging agent). This indicates that Bub1 is indeed required for the spindle checkpoint.

Because Bub1 contains a protein-kinase domain, Yu and co-workers tested whether Bub1 phosphorylates Cdc20, which is a key regulatory protein of APC/C, and found that it did. By contrast, a kinase-deficient form of Bub1 failed to phosphorylate Cdc20. When the authors next tested the effects of Bub1-mediated phosphorylation of Cdc20 on the ubiquitylation activity of APC/C^Cdc20, they found that substoichiometric amounts of Bub1 were sufficient to inhibit APC/C^Cdc20. Also, the kinase-defective Bub1 mutant failed to inhibit APC/C^Cdc20. Together, these observations indicate that Bub1 catalytically inhibits APC/C^Cdc20 by phosphorylating Cdc20.

Using mass spectrometry, the Yu team mapped the in vivo phosphorylation sites of endogenous Cdc20, which had been purified from nocodazole-treated cells. They identified six sites, all of which were also phosphorylated by Bub1 in vitro. That Bub1 phosphorylates Cdc20 in vivo was confirmed by checking the phosphorylation status of Cdc20 in Bub1-depleted cells — the slower-migrating, phosphorylated forms of Cdc20 that were present in wild-type, nocodazole-treated cells were absent in Bub1-depleted cells.

Next, the authors prepared a Cdc20 mutant (Cdc20^BPM) in which all six phosphorylation sites had been mutated. Bub1 failed to inhibit the ubiquitylation activity of APC/C^Cdc20BPM, which shows that Cdc20 — rather than APC/C itself — is required for the inhibition of APC/C^Cdc20.

Yu and colleagues isolated Bub1 from nocodazole-treated (metaphase-arrested) cells and from thymidine-treated (G1/S-arrested) cells. The kinase activity of Bub1 that was purified from the former cells was much higher than from the latter, which shows that the kinase activity is specifically activated after spindle-checkpoint activation. Overexpression of the non-phosphorylatable Cdc20^BPM mutant caused 40–50% of cells to arrest in mitosis, compared with >90% in the case of wild-type protein. So, elimination of the Bub1 phosphorylation of Cdc20 causes a substantial, yet partial, spindle-checkpoint defect. This is consistent with the existence of other APC/C-inhibitory mechanisms.

The spindle checkpoint is extremely sensitive, and Yu and colleagues propose that the catalytic mechanism for Bub1-mediated APC/C inhibition “...might be partially responsible for this remarkable sensitivity...”