Human cohesin dynamics and their regulation in M and S phase

Cohesin is a multimeric ring complex, which associates with DNA throughout most of the eukaryotic cell cycle. Its canonical role is to hold the two copies (sister chromatids) of each chromosome together from the time of their generation in S phase until their separation in M phase, thus ensuring the fidelity of mitosis. The complex mediates this so-called cohesion by topologically embracing the sister chromatids inside its ring structure composed of the three subunits Scc1, Smc1 and Smc3. For this thesis, I have studied cohesin's dynamic association with chromatin in M and S phase as well as its role at the mitotic centrosome, the main microtubule-organizing center of most eukaryotic cells. In order to separate sister chromatids in mitosis, cohesin must be removed from chromosomes. This is ultimately achieved by the action of the cystein protease separase, which cleaves cohesin at the centromeres (the primary constriction site of mitotic chromosomes) to release DNA. In vertebrates, however, the bulk of cohesin molecules, which is located on chromosome arms, is already removed from DNA in the first stage of mitosis, prophase, by a non-proteolytic mechanism called the prophase pathway. Since cohesin embraces the sister chromatids, prophase pathway signaling requires the ring to open up at at least one of its three subunit interaction sites ("gates"). However, which gate opens up for DNA exit has remained enigmatic. The same is true (at least in humans) for cohesin's DNA entry gate, which would be required to open up to topologically reload the ring complex after chromatid separation in telophase. For my thesis, I set out to identify cohesin's DNA exit (prophase) and entry (telophase) gates by employing the so-called FRB/FKBP system, allowing me to artificially close each of cohesin's gates individually in a conditional manner. I found that cohesin's DNA exit gate during prophase is composed of Smc3 and Scc1. To load the complex onto DNA, however, the gate situated between Smc1-Smc3 needs to open up. Utilizing different gates allows for more precise control over cohesin dynamics by maintaing the delicate balance between DNA entry and exit. Cohesin's association with chromatin remains very dynamic during G1 phase, but has to be stabilized as soon as the second sister chromatid is synthesized in S phase. Today, we know that this so-called cohesion establishment and DNA replication are two tightly co-regulated processes. However, how the replication machinery achieves DNA duplication, while simultaneously depositing the nascent chromatid inside the ring's lumen is still an unsolved problem. Using the aforementioned FRB/FKBP system in combination with an assay allowing me to assess DNA synthesis on a single replication fork level, I set out to determine whether the replisome might be able to pass through the closed ring or whether cohesin has to open up one of its three gates. These experiments revealed that cohesin gate opening unlikely to be required during replication but impaired dynamics of the complex in the preceeding G1 phase causes a dramatic reduction of replication fork velocities in the following S phase, likely by altering cohesin levels on chromatin. While these results confirm a strong correlation between cohesin dynamics and replication, they also argue for a model in which pre-S phase-loaded cohesin complexes represent the future cohesive fraction and, moreover, that during co-replicative cohesion establishment the replisome may pass through the closed cohesin ring. Finally, we and other groups have shown that cohesin plays a functional role at the centrosomes. Each centrosomes comprises two centrioles, which are rigidly coordinated ("engaged") in a perpendicular fashion. Separase-mediated proteolysis of centrosome- associated cohesin causes centriole disengagement, a prerequisite for centrosome duplication in the following S phase. In a collaboration with Lisa Mohr (University of Bayreuth, Germany), we found that the same factor, which protects centromeric cohesin from prophase pathway signaling until its proteolysis at the metaphase-to-anaphase transition, namely Sgo1, is also required to maintain centriole engagement. More specifically, human cells express a set of Sgo1 splice variants, which exclusively localize and function either at the centromere or the centrosome. Further studies revealed that Sgo1's determinant for centrosomal function lies in its C-terminal 40 amino acids, an area, which we therefore named the "centrosomal targeting signal of Sgo1" or CTS. Nonetheless, Sgo1's centromeric function is conserved at the centrosome, since my results demonstrate that CTS-containing Sgo1 variants protect centrosomal cohesin from the action of the prophase pathway by recruiting protein phosphatase 2 A (PP2A). Our results provide compelling evidence that the cell coordinates chromosome and centrosome cycles by multiple use of proteins like cohesin and its regulatory framework. However, employing specific Sgo1 splice variants allows these two processes to be precisely and individually controlled or maybe even uncoupled under certain circumstances as, for example, during spermatogenesis.

Level thesis:degree_level

thesis.doctoral

Grantor dc:publisher

Universität Bayreuth

Year

2015

Author dc:creator

• Buheitel, Johannes

Contributors dc:contributor

• Stemmann, Olaf

Repository record source_url

https://epub.uni-bayreuth.de/id/eprint/2499/

OAI identifier oai:identifier

oai:epub.uni-bayreuth.de:2499