Introduction
The maintenance of genome integrity during cell division impacts cell viability, speciation, birth defects and human disease. To maintain genome integrity, the cell must properly duplicate and segregate each chromosome without altering its integrity. Proper chromosome segregation requires specific chromosome structures that result from different types of higher-order chromosome organization. The exact biological functions of these chromosome structures and the molecular basis for their formation, maintenance and disassembly are major unsolved problems in cell biology. To maintain chromosome integrity, cells must suppress processes that lead to the formation of gross chromosomal rearrangements (GCRs) such as translocations, internal deletions or terminal truncations. Researchers have only begun to elucidate the features of chromosomes and the cellular factors that are critical for preventing GCRs. My laboratory uses budding yeast as a model to understand higher-order chromosome organization and chromosome integrity.

Chromosome Structure
Divergent eukaryotic Smc (structural maintenance of chromosomes) complexes mediate different types of higher order chromosome organization. These protein complexes include cohesin for sister chromatid cohesion, condensin for condensation, the dosage compensation complex for global transcription control, and Smc5/6 and MRX/N complexes for DNA repair. Therefore, Smc complexes provide a powerful platform to understand both the biological function and molecular basis for different types of higher order chromosome structure.

Using Smc mutants, my laboratory has shown that biological function of different types of higher order chromosome organization are highly interdependent. Different types of chromosome structure interact in at least three distinct ways to mediate proper chromosome morphogenesis in mitosis, meiosis and DNA repair. They cooperate to build other chromosomal structures like synaptonemal complex, regulate each other's assembly and finally regulate each other's disassembly.

Smc complexes are thought to organize chromosomes by tethering two different chromatin strands or two regions of the same strand. In principle such tethering could occur by dimerization of a canonical DNA binding protein. However, such a simple mechanism seemed unlikely after the discovery of the complex architecture of Smc complexes. We have studied one Smc complex, cohesin, in detail as a prototype to understand the molecular mechanism of tethering.

Cohesin tethers together the two replicated copies of a chromosome to facilitate chromosome segregation and to protect cells from DNA damage both at the site of the lesion and genome wide. We have shown that cohesins bind chromatin and then become cohesive. We also have shown that this conversion to the cohesive state is regulated by post-translational modifications including acetylation and phosphorylation. Our understanding of cohesin regulation allows us to lock cohesin in either the non-cohesive or cohesive-competent state, providing critical new tools to assess specificity of in vitro activities as well as to generate sufficient quantities of active complex for in depth biochemical and structure studies.

Chromosome Integrity
Dissecting the molecular basis for chromosome integrity is challenging because GCRs are usually rare and often either lethal or genetically silent. Factors that prevent GCRs have been identified in budding yeast from a candidate list using an elegant but labor-intensive genetic assay. While extremely important, these studies have likely missed processes that protect cells from GCRs because the candidate list included mostly DNA repair and replication factors. To identify novel processes that prevent GCRs, we developed a high throughput assay to follow GCRs using a yeast artificial chromosome (YAC). We used this assay to screen mutant yeast deleted for each of 5000 nonessential genes. Our studies identified novel mechanisms involving general mRNA metabolism and other metabolic pathways to maintain chromosome and genome integrity. We are currently investigating the molecular role of these processes in chromosome integrity.

Genome Evolution
Genome evolution often involves changes in chromosome number and large-scale changes in chromosome structure, exemplified by translocations and amplifications. The molecular basis for these changes in eukaryotes has largely been inferred from model systems that analyze events of a single cell division, minimizing the contribution of fitness/selection. In addition, large-scale changes in chromosome structure occur frequently through ectopic recombination between repetitive elements. Systems engineered with two or three repeat sequences have been used to gain information on the role of sequence divergence and different recombination pathways in ectopic recombination. However, it is not known whether this information is relevant to the more abundant and complex repeats in most genomes.

In collaboration with Yixian Zheng (Carnegie Institution and HHMI) and Maitreya Dunham (Princeton University) we have developed hybrid yeast as system to assess directly both the role of fitness and genome complexity in genome evolution. This hybrid results from a mating of two divergent Saccharomyces, cerevisiae with disperse repetitive retrotansposons (Ty) and bayanus that lacks these elements. We chose hybrid yeast because hybrids are an important intermediate in genome evolution during speciation. Their study provides an interesting complement to studies of purebred evolution already initiated in Maitreya's laboratory. Finally, hybrid yeast provides an excellent vehicle to study ectopic recombination at the genomic scale by having both a repetitive and divergent non-repetitive genome in the same organism/cell.

Our analyses of hybrids have provided significant insights into genome evolution. We show that long stretches of single stranded DNA form frequently near a double strand break in DNA as part of the mechanism to scan for homology during repair. This expanded homology search allows repeat sequences distal to the DSB to undergo ectopic recombination, expediting GCR and genome evolution. As a result of expanded homology search, the influence of repetitive DNA on genome evolution is much greater than its proportion of the genome. In addition, hybrids, that are grown under different nutrient limiting conditions for 100-300 generations, undergo repetitive-element-dependent translocations. Many translocations are the same as those in a natural cerevisiae-bayanus hybrid found in lager, validating this system to model the role of repetitive elements in genome evolution. In addition, evolved hybrid strains exhibit amplifications and loss of heterozygosity reminiscent of genome evolution in cancer cells. Thus hybrids are a model for other important aspects of genome evolution.