Search tips
Search criteria 


Logo of wtpaEurope PMCEurope PMC Funders GroupSubmit a Manuscript
Cell Cycle. Author manuscript; available in PMC 2010 October 19.
Published in final edited form as:
PMCID: PMC2957174

How Might DNA Enter the Cohesin Ring?

Cohesin is a gigantic ring-like complex, around 40 nm in diameter that mediates cohesion between two sister chromatids. Mounting evidence suggests that the interaction between cohesin complexes and DNA is topological with one or both sister chromatids entrapped within the cohesin ring—this is referred to as the ring model.1,2 However, a direct observation of the chromatin fibers within the cohesin ring is still missing. Alternative models of how cohesin might hold sister chromatids have been suggested, but compelling experimental evidence supporting these models is lacking.3 The cohesin complex consists of two SMC proteins, Smc1 and Smc3, and two non-SMC proteins, Scc1 and Scc3. Cleavage of the Scc1 subunit by a protease called separase opens the cohesin ring at the onset of anaphase. This releases cohesin from chromosomes and destroys sister chromatid cohesion. But how does cohesin load onto chromatin? If the ring model is correct, and cohesin entraps DNA topologically, then the cohesin ring must transiently open to allow DNA to enter the ring. Alternatively, cohesin rings may assemble de novo around the DNA. Gruber et al. now show that dissociation of two cohesin subunits is required for loading of cohesin onto DNA.4

The cohesin complex is a tripartite ring where Smc1 and Smc3 proteins are connected by their hinge domains on one side and the Scc1 protein closes the ring by connecting the Smc1 and Smc3 head domains on the other side. There are three possible gates where DNA might enter the cohesin ring. Either Scc1 dissociates from the Smc1 head or Scc1 dissociates from the Smc3 head or Smc1 and Smc3 hinge domains dissociate. Gruber et al. decided to test if artificial closing of these gates would prevent cohesin from loading onto DNA and generating sister chromatid cohesion. To prevent the cohesin ring from opening Gruber at al. elegantly linked two yeast cohesin subunits in a conditional manner by fusing them to human FKBP12 and Frb which efficiently dimerize in the presence of rapamycin. To their surprise, connection of Scc1 to SMC heads did not destroy cohesin function. In this respect, yeast cohesin seems to be different from its bacterial counterpart, because disengagement of the head domains is essential for a stable interaction between the bacterial SMC complex and DNA.5 However, different experimental setups preclude direct comparison of these results. If the ring model is correct, the only remaining possibility how DNA might enter the cohesin ring was by transient dissociation of SMC hinges. SMC hinges are highly conserved and have a very high affinity for each other.1 Could this tightly closed gate serve as a DNA entry point? To test this hypothesis, Gruber et al. artificially linked the Smc1 and Smc3 hinge domains in budding yeast S. cerevisiae. Connection of Smc1 and Smc3 hinges efficiently blocked association of cohesin with DNA and establishment of sister chromatid cohesion. This result suggests that SMC hinges are not merely dimerization domains, but they have an important role in association of cohesin with chromosomes, presumably serving as a DNA entry gate. This is consistent with studies describing DNA-binding properties of SMC hinge domains in vitro.5,6,7 Previous visualization of vertebrate cohesin hinges by electron microscopy showed no signs of open conformation.8 Interestingly, analysis of the crystal structure revealed both open and closed conformations of bacterial Smc hinges. At that time, only the closed conformation was considered to be biologically relevant.9

Gruber et al. make a novel and provocative claim that dissociation of SMC hinge domains is required for loading of the cohesin onto chromatin. This provides another important piece of evidence supporting the model that the interaction between cohesin and DNA is topological. It remains to be tested if this principle is unique to cohesin or applicable to all SMC proteins. A major challenge will be to find out if cohesion between sister chromatids mediated by cohesin is also topological. Previous studies suggested that ATP hydrolysis by SMC heads is required for a stable binding of the cohesin complex to chromatin.10,11 Does the ATP hydrolysis provide the energy for opening the hinge? If SMC hinges open to allow loading of the cohesin onto chromatin, there must be a mechanism ensuring reassociation of hinges after successful loading. Is this the role of the Scc1 subunit? Scc1 not only interconnects the Smc1 and Smc3 proteins, but it also regulates their ATPase activity.12

Further Reading

1. Haering CH, et al. Mol Cell. 2004;15:951–64. [PubMed]
2. Ivanov D, Nasmyth K. Cell. 2005;122:849–60. [PubMed]
3. Huang CE, Milutinovich M, Koshland D. Philos Trans R Soc Lond B Biol Sci. 2005;360:537–42. [PMC free article] [PubMed]
4. Gruber S, et al. Cell. 2006;127:523–37. [PubMed]
5. Hirano M, Hirano T. Mol Cell. 2006;21:175–86. [PubMed]
6. Chiu A, Revenkova E, Jessberg R. J Biol Chem. 2004;279:26233–42. [PubMed]
7. Yoshimura SH, et al. Curr Biol. 2002;12:508–13. [PubMed]
8. Anderson DE, et al. J Cell Biol. 2002;156:419–24. [PMC free article] [PubMed]
9. Haering CH, et al. Mol Cell. 2002;9:773–88. [PubMed]
10. Arumugam P, et al. Curr Biol. 2003;13:1941–53. [PubMed]
11. Weitzer S, et al. Curr Biol. 2003;13:1930–40. [PubMed]
12. Arumugam P, et al. Curr Biol. 2006;16:1998–2008. [PubMed]