 |
Cell Division, Cell Differentiation, and Genome Evolution
Summary: Yixian Zheng's lab studies how eukaryotic cells divide, differentiate, and evolve. Specifically, the lab studies how eukaryotic cells orchestrate their division and differentiation through the morphogenesis of the mitotic spindle in a number of systems, including the embryonic stem cells. The lab also uses hybrid yeasts from the Saccharomyces sensu stricto group to study eukaryotic genome evolution.
Spindle Morphogenesis
Proper spindle morphogenesis ensures the equal segregation of mitotic chromosomes and a regulated partitioning of other cellular components important for the survival, proliferation, and differentiation of daughter cells. Since mitosis involves drastic changes of cellular architecture, we believe spindle morphogenesis involves cellular components beyond the microtubule-based cytoskeleton and the chromosomes. We hope to explore whether and how the drastic changes in mitosis could offer the cell an opportunity to choose between proliferation, death, and differentiation in response to various signals.
One major structural component of the mitotic spindle is the microtubule cytoskeleton. A typical metazoan spindle mainly consists of three types of microtubules: astral microtubules, pole-to-pole microtubules, and kinetochore microtubules. The two astral arrays of microtubules are nucleated from the centrosomes at the spindle poles; these microtubules interact with the cell cortex and are required to position the spindle within the mitotic cell. The pole-to-pole microtubules make antiparallel interactions within the spindle and are required for keeping the two spindle poles apart. Finally, the kinetochore microtubules are involved in capturing chromosomes and regulating their congression and segregation. The assembly and organization of these microtubules found in the spindle is orchestrated by a large array of structural and regulatory proteins. Deciphering how these proteins function in a coordinated manner in mitosis represents a major challenge in understanding spindle morphogenesis.
Centrosome and Microtubule Nucleation
Centrosomes play important roles in mitotic spindle assembly by providing major microtubule nucleation and organization sites. Through the purification and study of the ~36S γ-tubulin ring complex (γTuRC), we showed that the complex is essential not only for microtubule nucleation at the centrosome but also for mitotic spindle assembly. We have identified the components of Drosophila (Dgrips 163, 128, 84, 91, 74, and Dgp71WD) and Xenopus (Xgrips 210, 110, and 109) γTuRC and have reconstituted a subcomplex of the γTuRC (which we termed γ-tubulin small complex [γTuSC]) that consists of two molecules of γ-tubulin and one molecule each of Dgrip84 and Dgrip91. Based on biochemical studies in my lab and structural studies in David Agard's lab (HHMI, University of California, San Francisco), a model of γTuRC has emerged (Figure 1).
Centrosome formation, involving the duplication of the pair of centrioles and the assembly of the surrounding pericentriolar material (PCM), is a poorly understood process. We utilized our PCM assembly assays to show that although γTuRC is essential for PCM assembly, additional activities are also required for both PCM assembly and maintenance. Using the PCM assembly assay, we biochemically enriched the activity required to assemble and maintain PCM and characterized one centrosomal protein that is required for the assembly and/or maintenance of γTuRC at centrosomes. To explore this active fraction fully, we identified ~360 proteins in this fraction in collaboration with John Yates (The Scripps Research Institute). Using an RNA interference (RNAi)-based phenotypic screen in Drosophila S2 cells, we identified a number of known and previously uncharacterized factors that regulate mitotic microtubule nucleation and spindle assembly. Characterization of the latter group of proteins should shed light on how centrosomes contribute to microtubule nucleation during both interphase and mitosis. (This research is supported in part by the National Institutes of Health.)
Spindle Assembly
Since our initial finding that the small GTPase Ran regulates multiple aspects of spindle assembly in mitosis, we have identified two important spindle assembly factors (SAFs), NuMA (a microtubule-binding protein) and XCTK2 (a microtubule-based motor), as downstream targets of Ran. NuMA and XCTK2 contain nuclear localization signals (NLSs) and are imported into the nucleus during interphase. During mitosis, the SAFs bind to importin-β through importin-α and become inactive in spindle assembly. Consequently, RanGTP is required to displace importin-αβ from the SAFs to stimulate spindle assembly.
Additional analyses have led us to uncover a novel Ran-signaling pathway that leads to activation of the essential mitotic kinase Aurora A. In mitosis, RanGTP releases TPX2—another SAF containing NLSs—from importin-α and -β and, which in turn allows TPX2 to bind and activate Aurora A kinase in the presence of microtubules (Figure 2). Structural studies by other groups have revealed that the enhanced binding of TPX2 to Aurora A creates an active kinase conformation ready for substrate binding. In addition, this active conformation protects Aurora A inactivation by phosphatase I.
Aurora A activation stimulated by RanGTP is an important regulatory mechanism in mitosis because Aurora A kinase regulates several aspects of spindle morphogenesis. To study how Aurora A might coordinately regulate microtubule nucleation and organization during spindle assembly, we coupled the kinase to magnetic beads of ~2.8 μm to make Aurora A–coated beads. Beads coated with Aurora A, but not its kinase-dead mutant, greatly stimulate the formation of both microtubule asters and bipolar spindles in Xenopus egg extracts in the presence of RanGTP. Although Aurora A itself does not have microtubule-nucleating activity, the beads coated with Aurora A remain associated with astral centers and spindle poles in this assay. These beads are able to recruit proteins required for both microtubule nucleation and organization during spindle assembly stimulated by RanGTP. This assay should facilitate the identification of proteins that are regulated by the Aurora A kinase in mitosis (Figure 3).
Aided by the bead-based assay, we uncovered a mitosis-specific function of lamin B, a type V intermediate filament protein. Lamin B is required for spindle morphogenesis, and it appears to function as a structural component of a membranous matrix that tethers a number of SAFs. This finding offers us an opportunity to explore the morphogenesis and partitioning of the membrane system in the context of spindle morphogenesis during cell division.
Since both cell cycle machinery and the Ran-signaling pathway regulate spindle morphogenesis, it is important to understand how the two pathways communicate with each other in mitosis. Our studies revealed that the cell cycle machinery regulates the Ran-signaling pathway by placing a high RanGTP concentration on the mitotic chromosome in mammalian cells. This local concentration of RanGTP provides a signal for spindle assembly toward the chromosomes. (This research is supported in part by the National Institutes of Health.)
Chromosome Segregation and Spindle Disassembly
The down-regulation of Cdk1 kinase, which leads to a cascade of phosphorylation/dephosphorylation events and protein degradation, is known to drive chromosome segregation and spindle disassembly. However, considering that spindle disassembly must be coordinated with many other morphological transformations, such as nuclear envelope re-formation and cytokinesis, regulatory pathways besides the cell cycle machinery must exist. Our lab is using Xenopus egg extracts to identify new regulators of mitosis. So far, we have discovered two such factors.
One of these regulators is p97-Ufd1-Npl4, a chaperone complex that recognizes ubiquitinated substrates on ER membranes in interphase. We found that p97-Ufd1-Npl4 also regulates spindle disassembly through a pathway that is separate from the mitotic exit network that drives Cdk1 inactivation. p97-Ufd1-Npl4 appears to regulate spindle disassembly by binding either directly or indirectly to a number of SAFs. This binding appears to sequester the SAFs from microtubules, which in turn allows the microtubule cytoskeleton to transform into the interphase network.
Our study of the second mitotic regulator—the de-ubiquitinating enzyme called FAM)—in conjunction with p97-Ufd1-Npl4 has demonstrated that ubiquitin regulates protein assembly dynamics in mitosis in a manner independent of protein degradation. Both FAM and the p97-Ufd1-Npl4 complex interact with Survivin, a subunit of the chromosome passenger complex required for chromosome segregation and cytokinesis. Whereas FAM removes the K63-ubiquitin linkage on Survivin in mitosis, p97-Ufd1-Npl4 is required for Survivin to acquire such linkages. A balanced K63 ubiquitination level on Survivin in turn regulates the dynamic binding of Survivin to centromeres, which is important for chromosome alignment and segregation.
Genome Evolution of the Hybrid Yeast
Genome duplication, which may involve individual genes, individual chromosomes, or a whole genome, is thought to be advantageous for evolutionary innovations. In plants, whole-genome duplication (WGD), either from a diploid genome or following interspecies hybridization, has long been recognized as a common phenomenon that drives the evolution of new species. Although WGD may be rare in animals, examples of speciation by WGD exist in some animal taxa such as amphibians and fish. Studies of gene paralogs in S. cerevisiae suggest that WGD also occurred in some fungi.
Although genome duplication is a widespread evolutionary phenomenon, little is known about the forces that drive the duplication or the molecular mechanism that underlies the genome reorganization/gene expression changes following duplication. We have begun to use the fully sequenced yeast sensu stricto species (S. cerevisiae, S. paradoxus, S. mikatae, and S. bayanus), which cross-hybridize naturally but produce mostly unviable gametes, to study genome evolution from the following angles: (1) Is the hybrid genome more likely to undergo whole-genome duplication than the diploid genome? (2) Is the hybrid genome more or less stable than the diploid genome under various environmental stresses? (3) Do hybrid and diploid genomes exhibit different reorganizations and expression patterns in response to various environmental stresses?
Last updated: December 21, 2006
|
|
|
|
