Nuclear pores set the speed limit for mitosis

The spindle assembly checkpoint prevents separation of sister chromatids until each kinetochore is attached to the mitotic spindle. Rodriguez-Bravo et al. (2014) report that the nuclear pore complex scaffolds spindle assembly checkpoint signaling in interphase, providing a store of inhibitory signals that limits the speed of the subsequent mitosis.

Mitosis requires a complex call-and-response of factors to ensure the accurate execution of events, including kinetochore-spindle attachment and anaphase onset. The anaphase-promoting complex/cyclosome (APC/C)advances mitotic progression by ubiquitinating cyclin B, while the spindle assembly checkpoint (SAC) proteins Mad1 and Mad2 transduce the signal from unpaired kinetochores into the formation of a mitotic checkpoint complex (MCC) that sequesters a key activator of the APC/C, Cdc20.When the activities of the APC/C and the SAC are in balance, mitosis proceeds efficiently with faithfully segregated chromosomes, leading to the formation of two euploid daughter cells. Our understanding of this antagonism between the APC/C and the SAC is now expanded by Rodriguez-Bravo et al. (2014) in this issue.

The authors initiate their study by revisiting the importance of Mad1 as a transducer of the SAC. It has been previously shown that Mad1 itself does not participate in MCCs, but catalyzes a conformational change in Mad2 to promote MCC assembly along with BubR1(Sironi et al., 2002). In addition, cells are more sensitive to RNAi-mediated depletion of Mad2 than Mad1, leading to controversy over the importance of Mad1 to mitotic regulation (Meraldi et al., 2004). In the present work, Rodriguez-Bravo et al. find that genetic ablation of the MAD1L1 locus causes the characteristic SAC defects of accelerated mitosis and chromosome missegregation, there by confirming that Mad1 is an essential early responder in the SAC.

In order to properly safeguard the genome, the SAC must prevent progression to anaphase in the presence of even one un-partnered kinetochore. However, several laboratories have reported the activity of SAC signaling outside this context. For instance, MCCs are found in cells in interphase, and even in cells lacking functional kinetochores (Fraschini et al., 2001; Maciejowski et al., 2010).To address this inconsistency, Rodriguez-Bravo et al. track cyclin B1 levels through mitosis in MAD1L1-null cells and find that cyclin B1 levels are decreased even before spindle formation. This finding points toward a role for Mad1 in inhibiting the APC/C before mitotic entry. What activates Mad1 in this context, and do MCCs formed in interphase also contribute to cell cycle regulation? It has long been known that Mad1 and Mad2 associate with nuclear pore complexes (NPCs) in interphase via direct interaction of Mad1 with the nuclear basket protein Tpr(Lee et al., 2008). This interaction could impact SAC function, as Tpr-depleted cells display accelerated mitoses and segregation errors (Schweizer et al. 2013; Rodriguez-Bravo et al., 2014). Importantly, Tpr is not among the group of nucleoporin proteins that decamp to the kinetochore upon NPC disassembly, so any role Tpr may have in mitotic regulation must be unfolding elsewhere.

The authors investigate the functional importance of this NPC tethering phenomenon and report that the Mad1-Tpr interaction leads to a similar outcome as the Mad1-kinetochore interaction: it activates Mad2, in turn catalyzing the formation of MCCs (Figure 1). While Mad2 can be activated at either the NPC or the kinetochore, the molecular details of these processes differ. Deletion of an N-terminal domain from Mad1 preserves kinetochore targeting and mitotic MCC formation but abolishes NPC targeting and interphase MCC production. Significantly, introducing this mutant into MAD1L1-null cells does not dampen mitotic progression or reduce rates of erroneous chromosome segregation. These data suggest that NPC-scaffolded interphase MCC production is essential for normal SAC function. To underscore this point, the authors directly tether truncated Mad1 to the NPC by fusing it to the NPC-targeting domain of Tpr. This chimera fully rescues all mitotic defects seen in both the MAD1L1-null and Tpr-deficient backgrounds. Whether kinetochore- and Tpr-generated MCCs are regulated similarly or function equivalently to limit APC/C activity are important questions for future work.

Figure 1. Nuclear pores impact the spindle assembly checkpoint.

(i) The Mad1-Mad2 complex is anchored to the NPC via Tpr in interphase, where it catalyzes the activation of O-Mad2 to C-Mad2 as it traffics through the nuclear pore. Activated C-Mad2 then accumulates in APC/C inhibitory MCCs in the cytosol. (ii) The Mad1-Mad2 complex is anchored to unoccupied kinetochores during mitosis, generating more MCCs and preventing the metaphase-to-anaphase transition until all chromosomes are bioriented on the spindle. NPC, nuclear pore complex; NE, nuclear envelope; MCC, mitotic checkpoint complex.

The interphase nucleus contains several thousand NPCs, while kinetochore-mediated MCC generation can be triggered by just one unpaired kinetochore. What are the relative contributions of these two pools to mitotic timing? Are kinetochore-generated MCCs merely a drop in the proverbial bucket? Analysis of MCC output by tracking cyclin stabilization indicates that mitosis duration can be tuned by gradations in MCC levels. That is, MCCs function as a rheostat, not a binary switch(Collin et al., 2013). Rodriguez-Bravo et al. propose that the Tpr-generated store of MCCs is essential for limiting APC/C activity at early stages of mitosis, and that abolishing this store could promote reckless, error-prone cell division by increasing the threshold of kinetochore-based signaling required to stall mitotic progression. The authors take their analysis one step further and generate a mathematical model that not only agrees with their data but also predicts that the checkpoint would be weakened when either NPC- or kinetochore-generated SAC signals were lost. In this context, it is relevant to consider effects of cancer-associated Tpr translocations (Köhler and Hurt, 2010) on SAC function by potentially decreasing the effectiveness of SAC signaling and predisposing cells to an euploidy, a direction that should be explored in the future.

For the checkpoint to respond to a single unoccupied kinetochore, it seems likely that kinetochore-derived SAC signals must be amplified in order to effectively stall mitosis. One possibility is that MCCs could template further Mad2 activation and additional rounds of MCC formation (Musacchio and Salmon, 2007). Another possibility for enhancing mitotic SAC signaling would be the continued activation of Mad1 by free Tpr during mitosis. The latter idea is supported by a report from the Cheeseman laboratory showing that the Tpr-Mad1 interaction persists in mitosis (Schweizer et al., 2013), although the data in the present study disagree in this regard.

Regulation of protein homeostasis is a recurrent motif in cell cycle regulation, which applies to the SAC as well, since Mad1 and Mad2 are susceptible to proteolysis when not bound to Tpr (Schweizer et al., 2013). In addition to preventing Cdc20 from activating the APC/C, MCCs also present Cdc20 as a substrate to other APC/Choloenzymes. Cdc20 is replaced as an APC/C cofactor in anaphase by Cdh1, which allows APC/C^Cdh1 to target a different set of substrates including Cdc20. This raises important questions such as whether the production of Tpr-scaffolded MCCs is coupled to the rise of Cdc20 levels through interphase, and how stable this stockpile of MCCs is. Overall, the findings here broaden our perspective regarding the importance of spindle assembly checkpoint signaling before the spindle is even present.