UC Berkeley

Cohesin is an essential protein complex required for chromosome dynamics and architecture. Its activities are required to promote sister chromatid cohesion, chromosome compaction, transcriptional regulation, as well as efficient DNA repair. Cohesin activity is regulated by several well-characterized protein accessory factors, but the molecular mechanism by which cohesin acts upon its substrate, DNA, remains elusive. In the absence of structural information elucidating how cohesin can interact with DNA to tether sister chromatids or mediate chromosome compaction, many models have been proposed. The simplest of these models postulates that cohesin forms a ring-like structure, and the topological entrapment of a sister chromatid is necessary and sufficient to mediate sister chromatid cohesion as well as condensation. Other models reject the notion that topological entrapment of DNA by cohesin is sufficient for sister chromatid cohesion or condensation, and propose that a second step after DNA binding is required for function.

In this dissertation, I utilized genetic and biochemical analyses to characterize how cohesin could tether DNA. Through mutagenesis of cohesin’s regulatory subunit, Mcd1p, I characterized a new domain in this key subunit that was required for viability, cohesion, and condensation. We utilized a new system for generating conditional null alleles in cohesin subunits, allowing for rapid analysis and characterization of these alleles in the absence of wild-type cohesin complexes. We expanded the conserved domain to a conserved ten amino acid domain in Mcd1p through the use of another directed mutagenesis screen in MCD1, and called this domain the Regulator of Cohesion and Condensation (ROCC). Importantly, mutant alleles in the ROCC box disrupted sister chromatid cohesion as well as chromosome condensation without perturbing cohesin’s stable binding to chromosomes. This observation was incompatible with the prevailing model in the field, where cohesin binding to DNA dictates its ability to generate cohesion.

While we had evidence that cohesin could stably bind DNA but unable to generate cohesion, my dissertation also uncovered an orthogonal line of evidence that further supported the possibility for higher order cohesin interactions. I previously characterized ROCC mutants as inviable in the absence of wild-type. However, when placed in trans with a second recessive inviable allele, this strain showed restoration of cohesin function, when none was expected. This inter-allelic complementation is consistent with a functional interaction between cohesin complexes, as evidenced by sister chromatid cohesion, chromosome condensation, and viability are all restored. Cohesin’s stable binding to chromosomes is restored in the trans configuration.

In sum, this dissertation furthers our understanding of how cohesin is able to tether sister chromatids on chromosomes. We show that cohesin can be stably bound to chromosomes, but unable to tether sisters, which contradicts the most common assumptions held in the field. Furthermore, we identify our evidence for interallelic complementation provides strong evidence for a physical communication between cohesin complexes.