Chromosomes are dynamic structures that are reorganized during the cell cycle to optimize them for distinct functions. Structural maintenance of chromosomes (SMC) and non-SMC condensin proteins associate into complexes that have been implicated in the process of chromosome condensation. The roles of the individual non-SMC subunits of the complex are poorly understood, and mutations in the CAP-G subunit have not been described in metazoans (Dej, 2004).

A role for Cap-G in chromosome condensation and cohesion has been demonstrated in Drosophila. The requirement of Cap-G for condensation during prophase and prometaphase is demonstrated; however, alternate mechanisms are demonstrated to ensure that replicated chromosomes are condensed prior to metaphase. In contrast, Cap-G is essential for chromosome condensation in metaphase of single, unreplicated sister chromatids, suggesting that there is an interplay between replicated chromatids and the condensin complex. In the Cap-G mutants, defects in sister-chromatid separation are also observed. Chromatid arms fail to resolve in prophase and are unable to separate at anaphase, whereas sister centromeres show aberrant separation in metaphase and successfully move to spindle poles at anaphase. A role for Cap-G during interphase in regulating heterochromatic gene expression is demonstrated (Dej, 2004).

Chromosomes undergo dynamic behaviors during mitosis to enable the precise separation of the two replicated sister chromatids. It is vital that the replicated sister chromatids are separated successfully. There are two crucial prerequisites for accurate segregation: (1) cohesion between the replicated chromatids must be maintained until anaphase and (2) compaction of the chromosomes into a manageable form, condensation, must be completed prior to metaphase. These processes require two major protein complexes, the cohesin and condensin complexes. Each of these complexes is founded upon a heterodimer of SMC proteins, which are chromosome-associated ATPases (Hirano, 1998; Hirano, 2002). Also within each complex are two or three non-SMC subunits, which contribute specific functions to the SMC holocomplex. Despite a similar structural paradigm, the condensin and cohesin complexes are functionally distinct. Although each complex was originally identified for unique functions during mitosis, it is now clear that both complexes are involved in a wide array of activities, including DNA repair, chromatid separation, and the regulation of gene expression (reviewed in Jessberger, 2002; Hagstrom, 2003; Legagnex: 2004; Dej, 2004 and references therein).

The structure and function of the cohesin complex is understood in the most detail and its structure has been elucidated (reviewed in Hirano 2000; Lee, 2001; Nasmyth 2002). The SMC subunits, SMC1 and SMC3, form two antiparallel coiled-coils (Hirano 2002). One of the two non-SMC subunits, SCC1/Mcd1/Rad21, associates the ends of the SMC coiled-coils into a ring structure. This ring structure holds the two sister chromatids together, perhaps by encircling them after S-phase. Cohesin is necessary for holding replicated sister chromatids together from S-phase until anaphase. The complex accumulates on chromosomes prior to S-phase and is maintained and activated through the process of replication. By the end of S-phase, replicated sister chromatids are associated through the cohesin complex at sites along the length of the arms. In yeast, the cohesin complex is maintained until anaphase along the chromosome. In metazoans, the bulk of the cohesin complex is displaced at prophase, but a subset of cohesin complexes is maintained at the centromere and perhaps other sites. This final population of cohesin complexes is lost at anaphase as the sisters separate (Dej, 2004).

The condensin complex, the subject of this biological overview, is a second SMC complex that is found in yeast and metazoans and is involved in chromatid segregation. It also contains two SMC subunits, SMC2 and SMC4 (Hirano 2002), and three non-SMC subunits, CAP-H, CAP-G, and CAP-D2 (Swedlow, 2003). These three subunits form an 11S regulatory subcomplex that is required to activate the SMC ATPases and to promote mitosis-specific chromatin binding of the holocomplex (Kimura, 2000). However, the individual functions of the non-SMC subunits within the complex remain undefined. Recent studies have identified another condensin complex containing alternate non-SMC subunits, CAP-G2, CAP-H2, and CAP-D3 (Ono, 2003). While there is a single condensin complex in both budding and fission yeast, condensin I and condensin II complexes are found in Xenopus and humans (Ono, 2003). Within the Drosophila genome, genes coding for a second CAP-H and a second CAP-D2 are found, but there appears to be only a single CAP-G protein (Ono, 2003; Dej, 2004 and references therein).

The condensin I complex was first identified biochemically in Xenopus extracts (Hirano, 1997). Sperm chromosomes in egg extracts depleted of condensin complex subunits assume a dispersed interphase organization. When the condensin complex is added back, the chromatin reorganizes into condensed chromosomes. This suggests a role in chromosome condensation supported by genetic analyses in yeast. In HeLa cells, depletion of condensin I or II complex subunits disrupt chromosome condensation, but depletion of subunits from both complexes has a more profound effect (Ono, 2003). Mutations in condensin subunits in yeast show precocious separation of sister chromatids in addition to defects in chromosome condensation (Saka, 1994; Strunnikov, 1995; Freeman, 2000; Ouspenski, 2000; Lavoie, 2002). Condensation defects in budding yeast were revealed through the use of fluorescent in situ hybridization (FISH) probes to rDNA, which appeared more dispersed in the mutants (Strunnikov, 1995; Freeman, 2000; Lavoie, 2002). In addition, FISH to euchromatic sites in fission yeast revealed loci to be more separated in condensin mutants (Saka, 1994) than in wild type (Dej, 2004).

In contrast, genetic analyses in metazoans to date have not delineated an essential role for the condensin complex in chromosome condensation. Embryonic lethal mutations in barren, the gene coding for the Drosophila homolog of CAP-H, show a failure to separate sister chromatids, but no described defect in condensation (Bhat, 1996). Animals with larval lethal mutations in gluon/smc4 also show defects in sister-chromatid separation. A partial effect on condensation is seen by an increase in chromosome width, but no change in the compaction along the length of the chromosomes (Steffenson, 2001). Further complicating the analysis of the role of condensin is the observation that in Drosophila S2 cells depleted of Barren by RNAi, chromosomes are poorly condensed with sister chromatids that are fuzzy and indistinct (Somma, 2003). Similarly, depletion of SMC4 by RNAi results in chromosomes that are undercondensed with sister chromatids that are unresolved (Coelho, 2003). In Caenorhabditis elegans, mutations in SMC4 show condensation defects at prometaphase, but little effect on condensation at metaphase and anaphase (Hagstrom, 2002). This is similar to observations in chicken cells lacking ScII/SMC2 in which chromosome condensation is delayed, but eventually reaches normal levels (Hudson, 2003). Together, these observations suggest that the condensin complex is not the only mechanism for compacting chromosomes in mitosis (Dej, 2004).

The genetic analysis of several mutations in the Cap-G gene was used to understand the role of Cap-G in Drosophila. Chromosome condensation was found to be compromised during mitosis in Cap-G mutant cells, but normal levels of condensation can be attained by metaphase. This suggests that there is a second pathway for condensing chromosomes that can compensate for a compromised condensin complex. Insight into this pathway comes from observations that, in the absence of replication, the Cap-G protein is required for chromosome condensation. In addition, in cells mutant for Cap-G, sister-chromatid arms are unable to resolve at prophase and sister chromatids show massive bridging defects at anaphase. While there is appropriate assembly of at least two centromere components, aberrant separation at the centromere is observed. Finally, it is shown that the Cap-G protein and perhaps the entire condensin complex may be required for chromatin-mediated gene expression in heterochromatic sequences (Dej, 2004).

These studies have demonstrated a role for Cap-G in chromosome condensation and cohesion in Drosophila. There is a requirement for Cap-G in the process of condensation during prophase and prometaphase; however, compensatory mechanisms ensure that chromosomes condense prior to metaphase. Reorganization of chromatin into condensed chromosomes is a process that involves the prior replication of chromatids and the condensin complex. Anaphase defects are also observed; specifically, sister-chromatid arms fail to separate. The separation defect is likely the result of the defect in sister-chromatid resolution during prophase that was evident in neuroblast mitotic squashes. In contrast, centromere separation is observed at anaphase and, in larval neuroblast preparations, this separation occurs aberrantly in metaphase figures. The role of Cap-G, and possibly of the condensin complex, is not limited to mitosis. Mutations have been identified that reveal roles for Cap-G during interphase in heterochromatic gene expression (Dej, 2004).

There is a striking condensation phenotype early in the cell cycle in Cap-G mutants. In prophase and prometaphase, condensation is nonuniform. By metaphase, condensation has achieved apparently normally levels, suggesting that a prolonged prometaphase enables chromosomes to achieve a high degree of condensation in the absence of a fully functional condensin complex. The role of the condensin complex in chromosome condensation prior to metaphase has been observed in other organisms. Mutations in the C. elegans smc-4 gene diminish chromosome compaction at prometaphase, but chromosomes are highly condensed by metaphase (Hagstrom, 2002). RNAi depletion of SMC4 in Drosophila S2 cells also shows aberrant condensation at prometaphase (Coelho, 2003). Similarly, in chicken cells lacking the condensin subunit ScII/SMC2, chromosome condensation is delayed, but chromosomes ultimately reach nearly normal levels of condensation (Hudson, 2003). Recently it has been shown that Xenopus and humans have two sets of condensin subunits (Ono, 2003). It is possible that in some organisms when one complex is unable to function the other can compensate partially and complete chromosome condensation by metaphase. However, while there are two CAP-H and CAP-D2 condensin subunits in Drosophila, there is only a single gene coding for a CAP-G subunit (Ono, 2003). This single Cap-G protein may be required in both complexes; therefore no such compensation is expected to occur in Drosophila. The significance of the alternate splice forms of the Cap-G gene is not known, although the two predicted proteins are similar across most of their lengths (Dej, 2004).

The prolonged prometaphase may be the result of activating the spindle checkpoint. This checkpoint may be used to monitor the degree of condensation at prometaphase to prevent sister-chromatid separation prior to complete condensation. The spindle checkpoint monitors the kinetochore-spindle attachment and delays anaphase until the appropriate bipolar connections are achieved and the chromosomes are congressed at the metaphase plate. It is thought that this checkpoint might monitor the tension at the kinetochores. It is possible that chromosome condensation might be monitored through this pathway, as a hypocondensed centromere and/or chromosome might reduce tension. In this way, the spindle checkpoint may delay progression through the cell cycle until the chromosomes are sufficiently condensed. The spindle checkpoint would require two sister chromatids for bipolar attachment, tension, and congression to occur (Dej, 2004).

The observation of a severe condensation defect at metaphase in other systems may be due to the absence of an active spindle checkpoint. For example, RNAi of Barren, the CAP-H homolog in Drosophila, demonstrated chromosome condensation defects at metaphase (Somma, 2003), but S2 cells have weak checkpoints controlling behavior in mitosis (Dej, 2004 and references therein).

Cap-G;double parked double mutants show that the condensin complex is dispensable for chromosome condensation by metaphase except in the absence of a replicated sister chromatid. What could replication provide to the condensation process? The cohesin complex could compensate for a faulty condensin complex, and replication is required to assemble the cohesin complex and establish cohesion. In this model the unreplicated sister chromatids in double parked mutants would contain inactive cohesin complex that would be unable to compensate for a faulty condensin complex in the process of chromosome condensation. Alternatively, experiments in Xenopus show that TopoII activity during replication is a prerequisite for setting up a structural axis required for the mitotic chromosome assembly (Dej, 2004).

Although there are no mutations in TopoII in Drosophila, some insights into whether TopoII could play the same role during replication in establishing a condensation-competent chromosome axis emerge from RNAi ablation in Drosophila cell culture. In S2 cells depleted of TopoII, mitotic chromosomes condense, but chromosomes are less compact at the metaphase plate (Chang, 2003). In these TopoII-depleted cells, Barren loading to centromeres and its dissociation at anaphase are normal and chromosome decondensation begins at anaphase (Chang, 2003). In SMC4-depleted S2 cells, TopoII localization is aberrant. In cells containing SMC4, TopoII appears in discrete regions along a defined chromatid axis, while in SMC4-depleted cells TopoII is associated diffusely with the chromosomes (Coelho, 2003). Thus, perhaps a condensation defect is seen at metaphase in the dup dcap-g double mutants because there was no replication, TopoII was unable to establish the appropriate architecture prior to mitosis, and the condensin complex is faulty during mitosis (Dej, 2004).

These studies present evidence that there is a distinct role for Cap-G in centromere segregation. The Cap-G, barren, gluon, as well as yeast and C. elegans condensin mutants exhibit the segregation of centromeres at anaphase, but a failure to separate arms. Aberrant separation of centromeres in Cap-G mutants were found at metaphase, prior to anaphase. Few other mutants show this defect in sister-chromatid centromere association. Mutations in cohesin subunits show premature separation of both the arms and the centromeres of chromatids (reviewed in Lee, 2001). In Drosophila, mutations in wings apart-like (wapl) cause aberrant separation of centromeres. However, in wapl mutants, the arms are resolved appropriately. This cytological observation suggests that there is a role for Cap-G and perhaps the condensin complex in mediating centromere association via heterochromatin. An effect on heterochromatic chromosome condensation in Cap-G mutants is further suggested by the observation that DAPI-bright repetitive sequences at the centromere appears to be expanded. A distinct role for Cap-G at the centromeres of the chromosomes is consistent with observations in other organisms that the condensin complex accumulates at centromeres (Dej, 2004 and references therein).

Consistent with the concept of a specific role for condensin in centromeric genomic regions is the observed role of condensin in regional gene regulation. Cap-G is required for the transcriptionally repressive state of centromere-proximal heterochromatin. Mutations in wapl not only show premature separation of centromeres at metaphase, but also, like Cap-G, act as dominant suppressors of variegation at the white locus. This emphasizes the intrinsic relationship between mitotic centromere structure and interphase heterochromatic organization. This is consistent with the widely held notion that the transcriptionally inactive state of mitotic condensation may be similar to the transcriptionally repressed heterochromatic regions of the genome. It is now becoming clear that the same proteins may establish both chromatic states (Dej, 2004).

Global gene repression in C. elegans is observed in XX hermaphrodites that downregulate gene expression from both X chromosomes. Dosage-compensation factors that resemble condensin subunits form a complex that associates with the chromosomes and mediate this chromosome-wide gene regulation. In C. elegans, a condensin complex containing MIX-1, SMC-4, and HCP-6 mediates mitotic chromosome condensation and a condensin-like complex containing MIX-1, DPY-26, DPY-27, and DPY-28 is required for dosage compensation (reviewed in Hagstrom, 2003). Silencing at the mating-type loci in Saccharomyces cerevisiae has also been found to require condensin subunits, specifically, CAP-D2 and SMC4, but not SMC2 (Bhalla, 2002). Perhaps in yeast, where there is a single condensin complex, a subset of condensin proteins assembles into a distinct condensin-like complex that is required for transcriptional silencing (Dej, 2004).

These studies reveal several distinct roles for the Cap-G condensin protein. In addition to its role in condensation and sister-arm resolution, these observations highlight the role of Cap-G and perhaps the condensin complex in centromere organization. This role is important for the association of sister chromatids during mitosis and for the regulation of heterochromatin-mediated gene expression during interphase (Dej, 2004).

GENE STRUCTURE

cDNA clone length - 4340 (isoform D).

Bases in 5' UTR - 195

Exons - 7

Bases in 3' UTR - 577

PROTEIN STRUCTURE

Amino Acids - 1155 amino acids (isoform D)

Structural Domains

There is a single Cap-G gene in Drosophila, but alternative splicing generates two forms of the protein that differ in their C termini. The distal splice form uses exons 5b and 6b and codes for a protein that is most homologous to the CAP-G subunits from other organisms. The alternative splice form uses exons 5a and 6a and codes for a protein with a C-terminal extension that is not homologous to regions within other CAP-G proteins (Dej, 2004). For information on the protein stucture see EMBL InterPro IPR000357 HEAT

date revised: 15 August 2005

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.