page 807 by Iouk et al. provides the first functional connection between kinetochores and nuclear pores through spindle checkpoint proteins and suggests that the two structures were linked long before vertebrates evolved.
Iouk et al. found budding yeast checkpoint proteins Mad1p and Mad2p at the nuclear pore complex (NPC). The Mad proteins interacted with a subcomplex of the NPC that contained, among other nucleoporins, Nup53p. The use of deletion strains indicated that Mad2p is linked to the NPC by its association with Mad1p, which in turn is bound to the Nup53p-containing complex. Unlike in vertebrates, which break down the nuclear envelope during mitosis, Mad1p remained at the NPC throughout the cell cycle. Mad1p is more than a passive observer at the pore, as nuclear import was slowed in yeast strains lacking Mad1p.
Upon checkpoint activation, Mad1p and Nup53p dissociated from certain neighboring nucleoporins, although they remained at the NPC. The rearrangements were unexpected, considering that the NPC remains intact during mitosis in yeast, and may be indicative of other events, including regulation of nuclear transport. Checkpoint activation also resulted in the phosphorylation of Mad1p and the release of Mad2p, which, unlike Mad1p, moved to kinetochores.
The authors speculate that the NPC may serve as a scaffold for the formation of certain checkpoint protein complexes. This function may be mediated by Mad1p, which is essential for checkpoint function, although it did not appear to leave the NPCs. Cells lacking Nup53p had reduced levels of Mad1p and Mad2p at the NPC, and very little Mad2p made its way to kinetochores. This did not prevent checkpoint activation, however, which is reminiscent of the recent demonstration in mammalian cells that the checkpoint can be activated with few or no Mad proteins at the kinetochore. ▪