SMC Family Proteins And Associated Factors In Mitotic Chromosome Segregation

Alexander V. Strunnikov, PhD, Head, Unit on Chromosome Structure and Function
Alexander Samoshkin, PhD, Visiting Fellow
Yoshimitsu Takahashi, PhD, Visiting Fellow
Stanimir Danailov Dulev, MSc, Graduate Student¹

Our main goal is to understand the biological roles and mechanisms of higher-order chromosome structure, which facilitates accurate chromosome segregation during cell division. The eukaryotic ATP-ases of the SMC family (Structural Maintenance of Chromosomes) form several essential protein complexes that determine higher-order chromosome structure and chromatin dynamics in eukaryotic cells. Currently, we focus primarily on mitotic chromosome condensation. The SMC complexes termed condensins play a pivotal role in chromosome condensation and are essential for chromosome segregation. Fission and budding yeast have only one condensin complex (condensin I) while multicellular organisms usually have two. The two condensin complexes both consist of five subunits: two shared SMC subunits SMC2/CAP-E and SMC4/CAP-C, which form the enzymatic (ATPase) and structural core of the complex, and three complex-specific non–SMC subunits. Condensins bind to DNA/chromatin and change its superhelical state, introducing positive DNA supercoiling. Our work has three major objectives: (1) characterizing regulatory pathways that target condensin to specific chromosomal sites in mitosis; (2) examining functional condensin’s requirements for distinctive chromosomal domains, such as nucleolar organizer and centromeres; and (3) quantitatively defining structural changes introduced into DNA during chromosome condensation in vivo. We identified several regulatory circuits and cellular requirements for condensin functions.

Regulation and specificity of condensin targeting to nucleolar chromatin

Takahashi, Dulev, Strunnikov

Characterization of the authentic binding sites for condensin activity in vivo is an essential step toward elucidating the molecular mechanisms of chromosome condensation. Our previous characterization of the S. cerevisiae rDNA (ribosomal RNA gene cluster) as a main target of mitotic condensin activity may have identified one of the major applications of condensin activity in all eukaryotes. To find out how condensin recognizes particular DNA sequences on a chromosome, we need to understand the molecular roles of post-translational condensin modifications.

Our previous work identified a promising candidate for a condensin regulator: Smt4p, which is the SUMO isopeptidase (Smt3p in budding yeast). New data suggest that balanced sumoylation regulates condensin; that is, both the addition and timely removal of SUMO moiety have specific effects on condensin targeting. To investigate whether Smt3p directly modifies condensin itself and/or some other proteins cooperating with condensin, we developed an unbiased screening procedure that permits the identification of the authentic Smt3p conjugates in vivo with differential-length SUMO tags. Using this “SUMO footprint” assay, we detected SUMO modification of four condensin subunits (Smc2p, Smc4p, Ycs4p, and Ycs5p). However, assigning a specific biological role to the modification proved to be a challenge because the mere discovery of the condensin subunits’ modification did not provide insight into the biological function of the modification. Eventually, we achieved a breakthrough by constructing GFP-Smt3p fusion expressed from the native SMT3 promoter and showing that SUMO conjugates are significantly enriched in the nucleolus—the essential hub of activities of condensin and both topoisomerases I and II. Further investigation of the biological role of condensin’s sumoylation led to the hypothesis that SUMO modification of condensin facilitates its targeting to the nucleolus, promoting condensin cooperativity with nucleolar pools of sumoylated topoisomerases I and II.

We tested our hypothesis with series of experiments. First, we mutated lysine residues to arginine (KR mutations) in all 16 of the sumoylation consensus sites (spread across the four condensin subunits) and assessed whether the generated SUMO-less condensin had any defects in its targeting to the nucleolus, chromosome condensation, chromosome segregation, and cell viability. Second, we combined the condensin KR mutants with topoisomerase mutations, which reduced the nucleolar presence of topoisomerases. Third, we identified the SUMO E3 enzyme for condensin—the Mms21 protein—and showed that Mms21p E3’s most significant activity is to modify condensin and cohesin complexes. We established that (1) SUMO is prominently enriched in the nucleolus; (2) triple SUMO E3 mutants are defective in rDNA segregation and maintenance; (3) genetic interactions among mms21-CH, top1, and top2 mutations signify a common/redundant pathway in rDNA maintenance; (4) condensin and cohesin binding to rDNA are directly regulated by Mms21-mediated sumoylation; and (5) condensin sumoylation is essential in the absence of Top1 and sumoylated Top2. These findings substantiated the paradigm that the SUMO “code” plays a role in the subnuclear compartmentalization of chromatin proteins, suggesting that SUMO-mediated co-targeting of individual proteins to a specific chromosomal compartment can induce the proteins’ functional cooperation.

Takahashi Y, Dulev S, Liu X, Hiller N, Zhao X, Strunnikov A. Cooperation of sumoylated chromosomal proteins in rDNA maintenance. PLoS Genet 2008;4:e1000215.
Takahashi Y, Iwase M, Strunnikov A, Kikuchi Y. Cytoplasmic sumoylation by PIAS -type Siz1-SUMO ligase. Cell Cycle 2008;7:1738-1744.
Takahashi Y, Strunnikov A. In vivo modeling of polysumoylation uncovers targeting of Topoisomerase II to the nucleolus via optimal level of SUMO modification. Chromosoma 2008;117:189-198.
Wang BD, Strunnikov A. Transcriptional homogenization of rDNA repeats in the episome-based nucleolus induces genome-wide changes in the chromosomal distribution of condensin. Plasmid 2008;59:45-53.

The role of active condensin in the functionality of the critical chromosomal sites—centromere and kinetochore

Samoshkin, Dulev; in collaboration with Cleveland

We previously conducted a ChIP-chip analysis of the genome-wide condensin binding pattern and showed that the peri-centromeric regions are enriched in condensin binding. Such enrichment suggested that condensin activity might facilitate centromere function. In addition, our earlier studies in budding yeast confirmed that condensin plays some role at centromeres; that is, ChIP-chip analysis revealed a notable increase in interphase enrichment of condensin near centromeres in mitosis. Furthermore, the mitotic checkpoint (spindle assembly checkpoint, or SAC) is the factor responsible for condensin mutant arrest, and both Dsn1 (a part of the MIND complex). We also found that Cse4 (the orthologue of the CENP-A centromeric histone) proteins were partially delocalized from centromeres in condensin mutants.

While our analyses uncovered a putative molecular interface between condensin and centromeres in yeast, they did not explain why checkpoints did not prevent chromosome non-disjunction and aneuploidy upon condensin dysfunction in humans. Therefore, we generated tools for efficient RNAi knockout of condensins in human cells and undertook biochemical, genetic, and cytological analyses of the structure of condensin-defective centromeres and kinetochores in human cells. The data showed that the centromere structure was abnormal and that the inter-kinetochore distance increased almost 2-fold in SMC2-depleted metaphase chromosomes. Furthermore, upon depletion of condensins I and II in both human and amphibian cells, the resulting cumulative centromeric structural defects activated the mitotic checkpoint; however, chromosome mis-segregation in anaphase still occurred. While we were investigating the molecular nature of such a slippage through the mitotic checkpoint in human cells, we found that depletion of condensin activity results in a significant loss of CENP-A from the centromere and a SAC-mediated metaphase delay. Given that CENP-A is loaded onto centromeres in the previous cell cycle, we analyzed, in collaboration with Don Cleveland, CENP-A enrichment during several cell divisions, using covalent fluorescent pulse-chase labeling (SNAP tagging).

The extreme stretching of centromeric chromatin by spindle forces led to the deformation of both inner kinetochores and the microtubule-capturing module (HEC1 complex), a configuration that strongly resembled a merotelic orientation. As a result, Aurora B was mis-localized and partially lost activity in condensin-depleted centromeres, suggesting that the correction of merotelic attachments before anaphase onset was severely impaired without condensin. Monastrol recovery experiments showed that the phenotype of the high incidence of merotelic attachment was nearly identical and epistatic in the cases of condensin depletion and Aurora B inactivation, indicating that condensin plays a pivotal role in the proper spatial positioning of the kinetochore relative to the microtubule-regulating protein complexes in the inner centromere (Figure 2.3).

Figure 2.3
Putative decondensation of centromeric chromatin resulting from the loss of condensin and CENP-A. The single sister centromere/kinetochore is shown. Left: wild-type condensed centromere attached to microtubules emanating from a single pole. Right side: condensin dysfunction. Once the centromeric chromatin (gray circles) becomes depleted and disorganized, the individual microtubule-attachment sites in the kinetochore (ovals) become semi-independent and acquire more flexibility to attach to a microtubule (truncated rectangles) for the incorrect pole. Dotted lines: DNA.

The results allow us to conclude that condensin dysfunction both compromises an important controlling mechanism (merotelic attachment correction) and leads to chromosome damage (breaks) and non-disjunction. Genome stability mechanisms are compromised in many cancer cell types and are thus believed to contribute to multi-step oncogenic transformations. The link of condensin to cancer was recently proposed for at least one cancer case. Our work shows that, in the instance of condensin dysfunction, chromosomal damage appears to result directly from centromere defects rather than from an independent non-disjoining defect at the chromosomal arms. Thus, a single hypomorphic condensin mutation can potentially generate a wide spectrum of genome instability signatures, as seen in cancers.

Dulev S, Aragon L, Strunnikov A. Unreplicated DNA in mitosis precludes condensin binding and chromosome condensation in S. cerevisiae. Front Biosci 2008;13:5838-5846.
Yong-Gonzalez V, Wang BD, Butylin P, Ouspenski I, Strunnikov A. Condensin function at centromere chromatin facilitates proper kinetochore tension and ensures correct mitotic segregation of sister chromatids. Genes Cells 2007;12:1075-1090.

¹Graduate Partnerships Program

Collaborators

Luis Aragón, PhD, MRC Clinical Sciences Centre, Imperial College School of Medicine, London, UK
Don W. Cleveland, PhD, Ludwig Institute for Cancer Research, University of California San Diego, La Jolla, CA
Mary Dasso, PhD, Program in Cellular Regulation and Metabolism, NICHD, Bethesda, MD
Louis Dye, BS, Microscopy and Imaging Core Facility, NICHD, Bethesda, MD
James McNally, PhD, Laboratory of Receptor Biology and Gene Expression, NCI, Bethesda, MD
Xiaolan Zhao, PhD, Memorial Sloan-Kettering Institute, New York, NY

For further information, contact strunnik@mail.nih.gov.