Solomon Lab research

The long range goal of our lab is to understand biochemically how cell growth and division are regulated by checkpoints within the cell and by controls imposed from the surrounding tissues. We are focusing on the regulation of the Cdc2 protein kinase, whose activity is required for entry into mitosis. There are many facets to Cdc2 activation--specific association with cyclin, multiple phosphorylations (both positive- and negative-acting), sensing of a threshold, and at least two feedback loops--that combine to produce the precise and abrupt activation of Cdc2 and transition into mitosis.

We are currently studying the kinases and phosphatases that control the Cdc2 phosphorylation state and how the changes in their activities are brought about and influenced by upstream signals. We are also studying the ubiquitin-mediated proteolysis of cyclins and related proteins at the end of mitosis and how this process is regulated by cellular checkpoints. It is hoped that combined approaches, ranging from genetics to biochemistry with pure components in organisms from yeast to man, will yield a more complete understanding of these macroscopic cellular behaviors.

Ongoing research projects

1. Mad3 Functions in the Spindle Assembly Checkpoint to Bind Cdc20 via Substrate Mimicry
2. Search for Novel Targets of the Anaphase Promoting Complex (APC)
3. Biochemical Comparison of Cdk2/Ringo A2 and Cdk2/Cyclin A: Cdc25 as a potential substrate of Cdk2/Ringo

Mad3 Functions in the Spindle Assembly Checkpoint to Bind Cdc20 via Substrate Mimicry
Janet L. Burton and Mark J. Solomon

The spindle assembly checkpoint monitors the attachment/tension between chromosome kinetochores and the kinetochore microtubules. Defects in the attachment/tension activate the checkpoint, which delays anaphase onset to prevent chromosome missegregation. The checkpoint delays anaphase by inhibiting the ubiquitin ligase activity of APCCdc20, which targets the anaphase inhibitor securin/Pds1p for ubiquitin-mediated proteolysis. The molecular players involved in the spindle checkpoint include Mad1, Mad2, Mad3, Bub1, Bub3 and Mps1. Both Mad2 and Mad3 are known to bind directly to Cdc20, but how this translates into inhibition of APCCdc20 activity has remained elusive. We and others have observed that Mad3 from many species contains a KEN box motif, a degradation signal often found in APCCdh1 substrates. We were therefore interested in investigating what role this motif may have in Mad3 function. We found that although the Mad3p KEN box plays no obvious role in Mad3p stability, it is nevertheless required for Mad3p checkpoint function. Unlike cells with wild-type Mad3p, cells containing Mad3p with a mutant KEN box (mkb) rapidly lose viability after brief exposure to microtubule disrupting agents and are able to grow in the presence of overexpressed Mps1p, indicating that they lack a functional spindle checkpoint. In agreement with this conclusion, Mad3-mkb cells (but not wild-type cells) continue to degrade Pds1p when challenged with spindle disrupting agents. Unlike wild-type Mad3p, Mad3-mkb is unable to bind Cdc20p in vivo, though it retains its ability to bind Bub3p, and displays reduced binding to recombinant Cdc20p in vitro. Interestingly, the APC-substrate Hsl1p can compete with Mad3p for Cdc20p binding in vitro in a D-box/KEN box dependent manner, suggesting that Cdc20p recognizes Mad3p using the same binding surface it uses for interacting with APC substrates. Taken together, these findings suggest that Cdc20p, like Cdh1p, recognizes the KEN box motif in vivo and raise the possibility that the checkpoint may function at least in part by antagonizing Cdc20p-substrate interactions. This finding may help unify diverse reports of how the spindle assembly checkpoint works by providing a common target in the Mad3-Cdc20 interaction. [top]

Search for Novel Targets of the Anaphase Promoting Complex (APC)
Denis Ostapenko and Mark Solomon

The anaphase promoting complex (APC) is a ubiquitin ligase that targets cell cycle proteins for proteasome mediated degradation. Although several APC substrates are known, such as cyclins and securin, additional proteins are likely to be targeted by the APC. To gain insight into cell cycle functions regulated by the APC, we set out to identify novel APC substrates in S. cerevisiae. By immunoblotting extracts from strains tagged for individual proteins, we identified several proteins that were present at lower level in G1 arrested cells than in cells arrested in metaphase, which is the pattern observed for authentic APC substrates. Among these candidate APC substrates was Iqg1, an IQGAP protein implicated in cell polarity and nuclear migration. We found that Iqg1 was rapidly degraded in wild type cells but stabilized in cdc23-1 cells, a conditional APC mutant, suggesting that APC activity was required for protein turnover. In addition, mutation of the APC activator Cdh1 prevented Iqg1 degradation. We have collaborated on studies of Iqg1 with John Pringle (Stanford University), whose lab has demonstrated that Iqg1 instability is linked to septin ring formation.

In a parallel set of experiments, we have characterized another potential APC substrate, Mps1, a multifunctional Ser/Thr protein kinase implicated in cell cycle regulation and the spindle assembly checkpoint. Mps1 protein levels fluctuated during the cell cycle, being low in G1 phase and high in mitosis. This fluctuation was APC dependent, as Mps1 levels increased in APC mutants. We identified potential degradation motifs (a destruction box and a KEN box) within Mps1 and demonstrated that mutations within these motifs prevented Mps1 degradation. We are currently assessing the phenotypes of cells expressing stabilized forms of Mps1 and are examining the role of Mps1 degradation in exit from the spindle checkpoint. [top]

Biochemical Comparison of Cdk2/Ringo A2 and Cdk2/Cyclin A: Cdc25 as a Potential Substrate of Cdk2/Ringo
Aiyang Cheng, Shannon Gerry, and Mark Solomon

Ringo, also called Speedy, belongs to a novel class of Cyclin-Dependent Kinase (CDK) activators that is essential for cell cycle transitions. We have identified a Speedy/Ringo-like gene in the most primitive branching clade of chordates (Ciona intestinalis), as well as four mammalian homologs. Of the mammalian proteins, two, Speedy/Ringo A and C, bind to Cdc2 and Cdk2, whereas Speedy/Ringo B binds preferentially to Cdc2. Despite their distinct CDK-binding preferences, both Speedy/Ringo A and B can promote the maturation of Xenopus oocytes and all three Speedy/Ringo proteins can bind to and activate CDKs in vivo. These mammalian Speedy/Ringo proteins exhibit distinct tissue expression patterns, though all three are enriched in testis, consistent with the initial observation that Xenopus Speedy/Ringo functions during meiosis. Human Speedy/Ringo A regulated G1/S progression in cultured cell lines. We recently found that human Speedy/Ringo C is required for the optimal growth of HEK293 cells. The existence of this growing family of CDK activators suggests that Speedy/Ringo proteins may play as complex a role in cell cycle control as the diverse family of cyclins. However, there are many important biochemical differences between how cyclin and Ringo activate Cdks.

We found that Cdk2-Ringo displays a broad substrate specificity, which is very different from the narrow consensus sites phosphorylated by Cdk2-cyclin A. Previous studies showed that CDK-cyclin preferred substrates containing (K/R) S/T P X (K/R) sequences. In striking contrast to Cdk2-cyclin A, we found that Cdk2-Ringo can tolerate almost any amino acid residue substitution at the +3 position. In addition, Cdk2-Ringo possesses low enzymatic activity toward conventional CDK substrates. For example, the overall kinase activity of CDK-Ringo toward histone H1 is as low as 0.08% of Cdk2-cyclin A in the standard histone H1 kinase assay. However, Cdk2-Ringo can phosphorylate non-conventional CDK substrates nearly as well as Cdk2-cyclin A, suggesting that Ringo may potentiate the activity of Cdk2 toward a select group of important but normally weak Cdk2 substrates. An additional difference between cyclin and Ringo is that Ringo can activate Cdk2 independent of the activating phosphorylation of Cdk2 on Thr-160. We compared the substrate specificity and enzymatic activity of Cdk2-Ringo in the presence or absence of the activating phosphorylation. For Cdk2-Ringo, neither the overall catalytic activity nor the substrate recognition required activating phosphorylation on Cdk2. In addition, CDK-Ringo is a poor substrate for the CDK-Activating Kinase (CAK), which phosphorylates Cdks on Thr-160. Therefore, the activation of Cdk2 by Ringo is CAK-independent.

We noticed that certain physiological substrates for CDKs accommodate non-canonical CDK phosphorylation motifs. One such protein is the Cdc25 dual-specificity phosphatase. Cdc25 controls CDK activity by removing inhibitory phosphorylations from CDKs, leading directly to their activation. CDKs can also phosphorylate Cdc25 at multiple sites, leading to further activation of Cdc25. There are three Cdc25 isoforms in humans and all of them appear to be phosphorylated by CDKs. Interestingly, among 32 (S/T)PXX sites in human Cdc25A, B, and C, only two sites fit the consensus CDK phosphorylation motif. We found that GST-Cdc25A, B, and C were phosphorylated by Cdk2-Ringo nearly as well as by Cdk2-cyclin A. In contrast, consensus site-containing Cdk2 substrates were phosphorylated about 1000-times as well by Cdk2-cyclin A as by Cdk2-Ringo. Phosphopeptide mapping suggested that Cdk2-Ringo and Cdk2-cyclin A phosphorylated Cdc25s on overlapping but not identical sites. These findings suggest that one way that Ringo could help jump-start the G1-S transition is via phosphorylation of weak Cdk2 substrates such as Cdc25. The small amount of active Cdc25 could activate Cdk2-cyclin A complexes, leading to further activation of Cdc25 and the concerted transition into S phase. [top]

research overview | publications | back to research projects | top