BIOLOGICAL OVERVIEW

Faithful segregation of the genome during mitosis requires that interphase chromatin be condensed into well-defined chromosomes. Chromosome condensation involves a multiprotein complex known as condensin that associates with chromatin early in prophase. The SMC4 subunit of condensin (SMC stands for structural maintenance of chromosomes) is encoded by the essential gluon locus in Drosophila. Drosophila SMC4 contains all the conserved domains present in other members of the structural-maintenance-of-chromosomes protein family. Drosophila SMC4 is both nuclear and cytoplasmic during interphase, concentrates on chromatin during prophase, and localizes to the axial chromosome core at metaphase and anaphase. During decondensation in telophase, most of the SMC4 leaves the chromosomes. An examination of gluon mutations indicates that SMC4 is required for chromosome condensation and segregation during different developmental stages. A detailed analysis of mitotic chromosome structure in mutant cells indicates that although the longitudinal axis can be shortened normally, sister chromatid resolution is strikingly disrupted. This phenotype then leads to severe chromosome segregation defects, chromosome breakage, and apoptosis. Thus SMC4 is critically important for the resolution of sister chromatids during mitosis prior to anaphase onset (Steffensen, 2001).

Chromosome condensation serves not only to package the long DNA molecules present in eukaryotic cells but also to organize the sister strands into chromatids that can be resolved from one another and from other chromosomes. Early work indicated that DNA topoisomerase II (topo II) is a structural component of mitotic chromosomes. While topo II activity is essential for establishing mitotic chromosome condensation both in vivo and in vitro, it may not be required to maintain the condensed state. New insights into chromosome dynamics came with the identification of a novel family of conserved chromosomal ATPases, the SMC proteins, reviewed in Cobbe (2000) and Hirano (1999). SMCs were initially identified in S. cerevisiae and in the analysis of the ScII protein that, like topo II, is a component of the chromosome scaffold fraction (Strunnikov, 1993). SMCs have been identified in all eukaryotic organisms examined to date and fall into six clearly discernible subfamilies: SMC1-4, Rad18, and Rad18 related. All SMC proteins are large polypeptides that share partial sequence conservation and similar structural organization. The N-terminal end contains a putative Walker A motif (ATP binding domain), and at the C-terminal end is located a characteristic 'DA' box (Walker B motif) involved in ATP hydrolysis. These domains are separated by a coiled-coil region, interrupted by a central globular hinge region. A complex containing a heterodimer of SMC2 and SMC4 has been shown to function in chromosome condensation (Hirano, 1994; Saka, 1994; Strunnikov, 1995; Sutani, 1999), while an analogous complex containing an SMC1/SMC3 heterodimer plays a role in sister chromatid cohesion (Steffensen, 2001 and references therein).

The 'condensin' complex, first identified in Xenopus egg extracts, is essential for mitosis-specific chromatin condensation in vitro (Hirano, 1994). XCAP-C and XCAP-E, the Xenopus SMC2/SMC4 homologs, along with three non-SMC proteins (XCAP-D2, XCAP-G, and XCAP-H) comprise a 13S condensin complex, while an 8S complex consists of XCAP-C and XCAP-E alone (not sufficient for mitotic chromosome condensation) (Hirano, 1997). A similar five-subunit complex has been identified in fission yeast, and mutations in the non-SMC subunits produce hypocondensed chromosomes that extend along an elongated spindle (Sutani, 1999). The Xenopus 13S condensin complex induces ATP-dependent positive supercoiling of a DNA template as a result of its stoichiometric binding to DNA (Kimura, 1997 and Kimura, 1999). The targeting of the complex and its supercoiling activity depend on CDK-dependent, mitosis-specific phosphorylation of XCAP-D2 and XCAP-H. The ATP-dependent positive DNA supercoiling that is associated with the condensin complex is proposed to be the driving force for chromatin condensation (Kimura, 1997). In the presence of topo II, condensin converts circular plasmid DNA preferentially into positive trefoil knots, and this observation implies that the condensin complex might generate an ordered array of positive solenoidal supercoils (Kimura, 1999). However, the mechanism by which the in vitro DNA supercoiling activity of 'condensin' may contribute to the overall condensation mechanism of protein-laden chromosomes is unknown (Steffensen, 2001 and references therein).

The advantages of combined genetic and cytological analyses make Drosophila an ideal system for studying the mechanics and regulation of mitosis during development. Of the condensin subunits, so far only the Drosophila homolog of XCAP-H (Barren) has been reported (Bhat, 1996). barren mutations exhibit defects in the segregation of chromosomes in embryonic neuronal precursor cells. The centromeres separate, but chromosome arms do not resolve, and they show chromatin bridges during anaphase. Examination of the cellular phenotype of barren mutants revealed a morphology similar to that exhibited by top2 (topo II) mutants of S. cerevisiae or cut3 or cut14 (SMC4 and 2, respectively) mutants of S. pombe (Saka, 1994). The observation that Barren and topo II coimmunoprecipitate and interact in a two-hybrid assay leads to the hypothesis that Barren might modulate topo II activity (Bhat, 1996). However, in budding yeast, the Barren homolog (BRN1) does not appear to be an essential activator of DNA topo II but is an important factor in the establishment and maintenance of chromosome condensation (Lavoie, 2000 and Ouspenski, 2000).

The Drosophila ortholog of the SMC4 subunit of condensin has now been cloned and characterized (Steffensen, 2001). Two P element insertion alleles have been characterized, corresponding to the previously described gluon locus (Kania, 1995), which cause lethality at different times in development. Mutations in gluon cause abnormal chromosome condensation, leading to chromosome breakage and missegregation during mitosis in both embryonic neuronal cells and larval neuroblasts. Detailed analysis of chromosome morphology in mutant cells suggests that SMC4 plays a crucial role in the final stages of mitotic condensation. Strikingly, in Drosophila SMC4 mutants, the longitudinal axis of a chromosome can be shortened, while width-wise chromatid compaction and sister chromatid resolution are severely disrupted (Steffensen, 2001).

Smc4/gluon encodes a protein that clearly belongs to the SMC4 subfamily of SMC proteins; this protein is one of the two SMC subunits of the condensin complex. In Xenopus and yeast, the protein is found in a multi-subunit complex consisting of SMC2/4 as well as CAP-G, -H, and -D2 proteins. Drosophila SMC4 is likely to be found in a similar complex in Drosophila: all subunit homologs have been identified by sequence homology, and Drosophila SMC2/SMC4/Barren and additional proteins can be coimmunoprecipitated from embryo extracts. SMC4 is an essential gene, since animals homozygous for either of two P insertion alleles die as embryos (glu¹) or as larvae/pupae. The maternal contribution of SMC4 is presumably sufficient to sustain embryos through the early syncytial blastoderm cycles. SMC4 was found to be expressed in all mitotically active tissues and developmental stages. The highest levels of protein are found in tissues containing a high number of proliferating cells, and this observation is consistent with SMC4 being required for both mitotic and meiotic cell divisions (Steffensen, 2001).

SMC4/Gluon exhibits a cell cycle distribution expected of a protein involved in the condensation and maintenance of mitotic chromosome structure. It is loaded onto condensing chromosomes early in prophase and dissociates late in anaphase/telophase when the condensation process is reversed. During these stages, the protein shows a restricted distribution running along the longitudinal axis of chromatids, similar to what has been observed for yeast topo II and ScII/SMC2 (Saitoh, 1994). A minor pool of Drosophila SMC4 remains associated with chromatin in telophase and interphase, whereas the major fraction becomes cytoplasmic at this time. The initial increase in chromatin association occurred before the onset of nuclear-envelope breakdown. While immunostaining for SMC4 has revealed interphase cytoplasmic localization in S. pombe (Sutani, 1999), in Xenopus the protein appears to be nuclear during interphase (Hirano, 1994). The human and chicken CAP-E [SMC2] proteins have also been reported to be nuclear. The differences in interphase localization could be due to differential solubility during fixation procedures of nuclear versus cytoplasmic SMC4. In Xenopus, mitosis specific phosphorylation of condensin controls its activity, whereas in S. pombe activity appears to be controlled by the shuttling of condensin in and out of the nucleus (Steffensen, 2001 and references therein).

Barren colocalizes with Drosophila SMC4 on condensed chromosomes and the localization of Barren is dependent on the presence of SMC4. These findings are consistent with participation in a common complex, as has been shown in Xenopus and yeast (Hirano, 1994 and Sutani, 1999). This complex may only be formed late in interphase; Barren and SMC4 do not appear to colocalize in the cytoplasm. This finding is consistent with the presence of a separate condensin 'regulatory' complex containing the non-SMC subunits (Kimura, 2000). On condensed chromosomes, Drosophila SMC4 and Barren are more concentrated in the centromeric region. This observation is consistent with chromosome condensation initiating at the centromere and then spreading distally, similarly to the way in which mitosis-specific phosphorylated forms of histone H3 accumulate. A role for SMC4 and Barren in the condensation of the centromeric heterochromatin underlying the kinetochore may be necessary for the proper assembly of the kinetochore. Indeed studies in S. cerevisiae have implicated BRN1 in normal kinetochore function (Ouspenski, 2000). Interestingly, topo II also shows a similar enrichment at centromeres in a number of mammalian cells (Steffensen, 2001 and references therein).

Drosophila SMC4 is required for sister chromatid resolution. To analyze in detail the effects of SMC4 mutants on chromosome morphology, two aspects of chromosome condensation were quantified in neuroblasts. The shortening of the longitudinal axis of the chromosome is unaffected in gluon/SMC4 mutants and occurs with the same kinetics as in wild-type cells. However, a significant reduction was found in the ability of SMC4 mutant cells to resolve sister chromatids during prometaphase. Arresting mutant neuroblasts with colchicine to allow more time for the condensation process does not rescue the defect in resolution, while chromosome shortening proceeds as in wild-type cells (Steffensen, 2001).

How is the chromosome axis shortened when Drosophila SMC4 levels are reduced? The axial shortening of chromosomes observed in mutant cells suggests that either the residual level of SMC4 is sufficient to support this aspect of condensation or other factors are responsible for this process. The fact that the kinetics of length reduction appear to be independent of SMC4 levels would argue against SMC4 involvement in this process. Several results support the hypothesis that topo II might be the enzyme responsible in part for the shortening of the axes: (1) topo II has been shown to be required for the hypercondensation of mitotic chromosomes after the treatment of yeast cells with microtubule-depolymerizing agents; (2) blocking mammalian cells with the ICRF topo II inhibitors leads to hypocondensed mitotic chromosomes; (3) topo II is targeted to chromatin independent of the condensin complex in vitro. Alternatively, other members of the condensin complex may still mediate arm shortening (Steffensen, 2001 and references therein).

The fact that Drosophila SMC4 mutant cells are able to shorten their longitudinal chromosome axes but are unable to properly resolve sister chromatids indicates that these two aspects of chromosome condensation can be separable by mutation. Current models suggest that one purpose of chromosome condensation (probably shortening of the axis) is to displace the topo II catenation/decatenation equilibrium in the direction of decatenation and thereby promote the resolution of DNA catenanes produced during replication. In this context, the data suggest that wild-type levels of SMC4 are required to drive topo II toward decatenation and thereby promote sister chromatid resolution. However, the failure of sister chromatid resolution at reduced SMC4 levels could also be due to the inappropriate maintenance of the cohesion of sister chromatid arms. This interpretation is supported by recent results indicating that cohesion along chromatid arms is removed during early stages of chromosome condensation. The fate of Drosophila cohesin components in cells that carry mutations of condensin components will be an important question to address (Steffensen, 2001 and references therein).

Mutations in Drosophila SMC4 cause abnormal chromosome segregation. SMC4 mutations affect not only chromosome condensation but also chromosome segregation. All alleles examined in this study cause chromosome segregation defects, manifest as chromatin bridges in anaphase and telophase. Chromosome segregation defects are likely to be the consequence of aberrant sister chromatid resolution. Studies in yeast of mutations for top2, cut3, and cut14 strongly suggest that proper chromosome condensation is essential for normal sister chromatid separation. The chromatin bridges present in Drosophila SMC4 mutant cells are unlikely to be resolved, as significant chromosome breakage can be observed. The ultimate consequence of this state is probably the increased level of apoptosis observed in mutant brains (Steffensen, 2001 and references therein).

PROTEIN STRUCTURE

Amino Acids - 1409

Structural Domains

A cDNA has been cloned containing an open reading frame (ORF) encoding a protein with a predicted molecular weight of 160 kDa. The predicted structure of the protein closely resembles that described for other SMC proteins. Phylogenetic analysis of the protein sequence shows that it belongs to the SMC4-type protein family. The Drosophila protein is closely related to other SMC4-type proteins, including Xenopus XCAP-C (41% identity), human HCAP-C (39% identity), S. cerevisiae SMC4 (35% identity), and S. pombe CUT3 (34% identity) orthologs, and it is more distantly related to other SMC-type proteins. Accordingly, the gene has been named Drosophila SMC4 (Steffensen, 2001).

date revised: 15 March 2001

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.