New & Noteworthy
July 2, 2013
When Nik Wallenda recently made his incredible tightrope walk over a 1500 foot-deep gorge, the attachment of the cable he walked on was critical. If that had failed, it would have been a very unhappy ending for Nik.
Something equally dramatic can happen in a cell. If the attachment of spindle microtubules to chromosomes during cell division fails, then the chromosomes don’t end up in the right place. When this happens, the cell can end up dead, or even worse, cancerous. This is as bad as falling off a tightrope without a net!
In a cell, the chromosome is attached to the spindle with something called the kinetochore. It is like the spike driven into the side of the gorge the tightrope walker is going over. One end is attached to the chromosome (the side of the gorge) and the other is attached to the spindle (the rope that is tied to the spike).
This is where the analogy ends though…a kinetochore is way more complicated than a metal spike. It is a huge, multi-protein complex with lots of specialized parts. The way in which the whole complex assembles still isn’t completely understood.
In a new paper in GENETICS, Akiyoshi and coworkers unraveled a bit of the mystery behind it. They found that phosphorylation by a highly conserved protein kinase known as Aurora B (Ipl1p in S. cerevisiae) of one kinetochore subunit, Dsn1p, provides some of the glue that holds the structure together. More specifically, they found that phosphorylated Dsn1p does a better job at keeping inner kinetochore proteins attached to the complex. It drives the spike deeper into the gorge.
The researchers mutated two residues in Dsn1p that are sites for Ipl1p phosphorylation. They mutated one or both to alanine, which prevents phosphorylation, or to aspartic acid, which mimics the phosphorylated state. They found that preventing phosphorylation of these sites loosened the complex and keeping them “phosphorylated” tightened it.
First, to try to look at what happens when Dsn1p isn’t phosphorylated by Ipl1p, they mutated the two sites to alanine. Either site could be mutated with no apparent effects, but mutating both was lethal. Clearly these sites are doing something!
The researchers got around this lethality issue by mutating a third site in Dsn1p. This site is a target for phosphorylation by a different kinase, Cdk kinase (Cdc28p). The idea is that preventing phosphorylation by Ipl1p makes Dsn1p unstable, but then preventing phosphorylation by Cdc28p can stabilize the mutant protein.
Now that they had a living yeast strain in which Dsn1p wasn’t phosphorylated by Ipl1p, they could look to see what was different about the kinetochore in this mutant. When they pulled down the mutant Dsn1p using antibody and a Flag-tag, it brought down normal levels of outer kinetochore proteins but reduced levels of inner kinetochore proteins. So this suggested that Ipl1p phosphorylation promotes interactions between Dsn1p and inner kinetochore proteins.
Supporting this idea, an ipl1 mutant that phosphorylated Dsn1p to a lower extent showed lower-than-wild-type levels of inner kinetochore proteins associated with Dsn1p. And, when they looked at a mutant where those Dsn1p residues were changed to aspartic acid, mimicking constant phosphorylation, higher levels of inner kinetochore proteins were pulled down. All of this evidence, and more, points to Ipl1p phosphorylation of Dsn1p as critical for attachment of inner kinetochore proteins to the kinetochore complex.
In yeast there is just one Aurora kinase, and Dsn1p is just one of its substrates. In human cells there are multiple versions of Aurora, and they are implicated in cancer development. Clearly, yeast will be a helpful model in understanding all the details of how Aurora influences kinetochore structure and chromosome segregation. And that will be a much more impressive and useful feat than a tightrope walk!