Nup53p is a target of two mitotic kinases, Cdk1p and Hrr25p.

Nuclear pore complexes (NPCs) form channels across the nuclear envelope and provide the sole sites of molecular exchange between the cytoplasm and nucleoplasm. The NPC is a target of a number of post-translational modifications, including phosphorylation, yet the functions of these modifications are ill defined. Here, we have investigated the mitotic specific phosphorylation of a yeast nucleoporin Nup53p. Two kinases were identified that phosphorylate Nup53p: the mitotic kinase Cdk1p/Cdc2p/Cdc28p and the casein kinase Hrr25p. Hrr25p was identified by screening 119 yeast kinases for their ability to phosphorylate Nup53p in vitro. Conditional alleles of Hrr25p support the conclusion that Hrr25p phosphorylates Nup53p in vivo. We further demonstrated using solution binding and affinity purification assays, that Hrr25p directly binds Nup53p in an interaction that is destabilized by the phosphorylation of Nup53p. Consistent with this observation, we observed that Hrr25p moves between distinct locations in the cell during the cell cycle including the nucleus, the cortex of the emerging bud and the spindle pole bodies. Cdk1p also contributes to Nup53p phosphorylation as specific inhibition of Cdk1p or mutation of Cdk1p consensus sites partially blocked its phosphorylation. The ability of nup53 alleles containing Cdk1p site mutations to complement synthetic defects of nup53 Delta nup170 Delta strains is linked to a function for Nup53p in the spindle assembly checkpoint.

Cyclin A degradation employs preferentially used lysines and a cyclin box function other than Cdk1 binding.

Cyclin A is targeted for mitotic destruction by the anaphase promoting complex/cyclosome (APC/C) and degradation proceeds even when proteolysis of other APC/C substrates are blocked by the spindle assembly checkpoint. Instead of a simple destruction box, a complex N-terminal destruction signal has been implicated in Cyclin A. We show here that Drosophila Cyclin A destruction employs both N- and C-terminal residues, which emphasize that a synergistic action by different parts of the protein facilitates recognition and degradation. The first KEN box, first D-box and an aspartic acid at position 70 are required at the N-terminus and they make additive contributions when the spindle checkpoint is active. From the C-terminal region, the cyclin box contributes. Single point mutations in these four elements abolish mitotic destruction. Additionally, eight lysines in the neighborhood of the N-terminal signals, which could serve as potential ubiquitin acceptor sites, are preferentially used for proteolysis. Mutations in these lysines and the N-terminal signals cause mitotic stability. However, mutating the lysines alone, only delays mitotic progression. Thus, presumably, lysines elsewhere in the protein are used when the preferred ones are absent and this requires the N-terminal signals. Furthermore, our results suggest that some function of the cyclin box other than Cdk1 binding promotes spindle checkpoint-independent recognition of Cyclin A by the APC/C.

Requirements of cyclin a for mitosis are independent of its subcellular localization.

Cyclin A (CycA), the only essential mitotic cyclin in Drosophila, is cytoplasmic during interphase and accumulates in the nucleus during prophase. We show that interphase localization is mediated by Leptomycin B (LMB)-sensitive nuclear export. This is a feature shared with human CyclinB1, and it is assumed that nuclear accumulation is necessary for mitotic entry. Here, we tested if the unique mitotic function of CycA requires nuclear accumulation. We fused subcellular localization signals to CycA and tested their mitotic capability. Surprisingly, nuclear accumulation was not required, and even a membrane-tethered form of CycA was able to induce mitosis. We noted that Cyclin B (CycB) protein disappears prematurely in CycA mutants, reminiscent of rca1 mutants. Rca1 is an inhibitor of Fizzy-related-APC/C activity, and in rca1 mutants, mitotic cyclins are degraded in G2 of the 16(th) embryonic cell cycle. Overexpression of Rca1 can restore mitosis in CycA mutants, indicating that the mitotic failure of CycA mutants is caused by premature activation of the APC/C. The essential mitotic function of CycA is therefore not the activation of numerous mitotic substrates by Cdk1-dependent phosphorylation. Rather, CycA-dependent kinase activity is required to inhibit one inhibitor of mitosis, the Fzr protein.