G Shirleen Roeder

Eugene Higgins Professor Emeritus of Molecular, Cellular, and Developmental Biology and Professor of Genetics; Investigator, Howard Hughes Medical Institute

Research Interests

Meiotic Chromosome Behavior; Cell Cycle Checkpoints; Meiotic Recombination; Synaptonemal Complex

Research Summary

Our research is focused on using budding yeast to study chromosome pairing, synaptonemal complex formation and genetic recombination during meiosis, and the coordination of these processes with meiotic cell cycle progression.

Extensive Research Description

Our research is focused on the structure and behavior of meiotic chromosomes, using Saccharomyces cerevisiae as a model system. Specifically, we are interested in the interhomolog interactions that occur during meiotic prophase and are necessary for reductional chromosome segregation at the first meiotic division. Important events include homologous chromosome pairing, assembly of the synaptonemal complex (SC), genetic recombination, and the formation of chiasmata.

SC Proteins

The SC is an elaborate proteinaceous structure that assembles along the lengths of paired chromosomes during meiotic prophase. The Zip1 protein is a major building block of the complex, where it serves as a bridge between the cores of homologous chromosomes.

A complex of proteins, called the synapsis initiation complex (SIC), promotes the polymerization of Zip1 along chromosomes. Components of this complex include Zip2, Zip3 and Zip4. In the zip2 and zip4 mutants, there is a complete failure of SC formation; in contrast, a subset of synapsis initiation events does occur in the absence of the Zip3 protein.

Recent studies have revealed novel and unexpected functions for the Zip1 protein. In addition to its role as a structural component of the SC, Zip1 couples chromosomes at their centromeres, represses crossing over near centromeres, and promotes the segregation of nonexchange chromosomes. Furthermore, Zip1’s special affinity for centromeres defines these chromosomal regions as the earliest sites of synapsis initiation. Importantly, synapsis initiation at centromeres is subject to special regulatory mechanisms that prevent SC formation in the absence of meiotic recombination.

Zip1 Promotes Centromere Coupling

Studies of the spo11 mutant, which fails to initiate meiotic recombination or make SC, revealed an unexpected role for the Zip1 protein. In the spo11 mutant, Zip1 localizes to foci on chromosomes, and these foci coincide with centromeres. Furthermore, there are 16 foci per nucleus, representing 16 centromere "couples". (A diploid yeast cell contains 32 chromosomes.) In the spo11 mutant, coupling occurs predominantly between the centromeres of nonhomologous chromosomes and is dependent on Zip1 and independent of SICs.

In wild-type cells, centromeres are coupled from the earliest stages of meiotic prophase. Prior to the initiation of SC formation, most centromere couples involve nonhomologous chromosomes, but couples become increasingly homologous as meiosis progresses. Thus, centromere couples must switch partners until the centromeres of homologous chromosomes become coupled to each other. We propose that centromere coupling facilitates homolog pairing by holding two chromosomes together in a stable configuration while homology is being assessed.

Zip1 Promotes Synapsis Initiation at Centromeres

Centromeres also play a role in synapsis initiation. At an early stage in synapsis, when short linear stretches of Zip1 are first observed, about 80 percent of these regions of Zip1 staining are associated with centromeres. Furthermore, although SICs are not found at centromeres in spo11 nuclei, they are found – transiently – at centromeres in wild type. Thus, at early times in meiotic prophase, most synapsis initiates at centromeres. As meiosis progresses, synapsis initiation at noncentromeric locations makes an increasingly important contribution.

Studies of synapsis initiation at centromeres have revealed a number of unexpected features of the synapsis process. First, synapsis is often unidirectional, with Zip1 extending outward in only one direction from the centromere. Second, SICs are not fixed in location; instead, they move at the leading edge of Zip1 polymerization. These observations contrast with the traditional view that SICs remain fixed in position and synapsis extends outward in both directions from each SIC.

It remains to be determined whether synapsis initiation at noncentromeric locations is mechanistically similar to synapsis initiation at centromeres, with respect to directionality and SIC movement. There is at least one important difference between the two categories of synapsis initiation. Synapsis initiation at centromeres is independent of the Zip3 protein, whereas synapsis initiation at noncentromeric locations requires Zip3.

Zip1 and Fpr3 Prevent Synapsis Initiation in the Absence of Meiotic Recombination

Zip1 couples the centromeres of nonhomologous chromosomes in a spo11 mutant and at early stages of meiotic prophase in wild type. What prevents Zip1 from polymerizing along the lengths of chromosomes under these conditions? Synapsis can take place between nonhomologous chromosomes under certain conditions, so there must be a regulatory mechanism that normally prevents SC formation until homologs are paired and recombination is initiated. To characterize this regulatory network, we looked for mutations that would allow extensive Zip1 polymerization in a spo11 mutant background. We found that a mutation in the FPR3 gene, in conjunction with a mutation in the ZIP3 gene, allows synapsis to occur even when recombination is not initiated. Most of the synapsis that takes place in the spo11 fpr3 zip3 triple mutant involves nonhomologous chromosomes, and almost all synapsis initiates at centromeres. Surprisingly, the synapsis that occurs in the absence of the Fpr3 and Zip3 proteins is independent of Zip2, which previously has been shown to be absolutely required for SC formation.

Taken together, these data suggest that Fpr3 and Zip3 act in parallel and redundant pathways to prevent synapsis in the absence of double-strand breaks. The Zip2 protein may function specifically to overcome the inhibition imposed by Zip3 and Fpr3. We postulate that Fpr3 and Zip3 act by modifying Zip1 and/or other SC proteins. Zip3 is a SUMO ligase; Fpr3 is a proline isomerase. It is surprising that Zip3 performs different functions at different locations. Zip3 inhibits synapsis at centromeres, whereas it promotes synapsis at noncentromeric locations.

Zip1 Represses Crossing Over Near Centromeres

Meiotic recombination events, specifically crossovers, are nonrandomly distributed along chromosomes. One aspect of this nonrandomness is that crossing over is repressed near centromeres. We have measured crossing over on a genome-wide basis using DNA microarrays to monitor the segregation of ~8,000 polymorphisms. These results reveal a striking (~10 fold) decrease in crossing in the 10-kbp intervals on both sides of the centromere, as predicted by genetic studies. Furthermore, the centromeric repression of crossing over is abolished by a zip1 null mutation. The zip1 mutation also increases the rate of gene conversion in centromere-proximal intervals, but it does not affect the frequency of meiotic double-strand breaks. Thus, Zip1 most likely affects crossing over near centromeres by promoting interactions between sister chromatids at the expense of interhomolog interactions.

Crossovers that occur too close to centromeres cause chromosomes to missegregate in meiosis. Thus, Zip1 may help to ensure proper chromosome segregation by preventing centromere-proximal crossovers.

Zip1 Promotes Distributive Disjunction

Crossovers promote the correct segregation of chromosomes at the meiosis I division by promoting the formation of chromatin bridges called chiasmata. However, if one or a small number of chromosome pairs fail to crossover, there is a backup system that facilitates their proper segregation. In wild-type cells, the rate of nondisjuction for nonexchange chromosomes is ~10%. This rate is elevated two to three fold by a zip1 mutation, suggesting a role for Zip1 in promoting the segregation of nonexchange chromosomes. The effect of zip1 on nonexchange chromosome segregation is not due indirectly to its reduced level of crossing over and a consequent increase in the number of nonexchange chromosomes: an msh4mutation decreases crossing over to the same extent as Zip1 but does not increase the rate with which nonexchange chromosomes missegregate. Furthermore, cytological data indicate that the Zip1 protein holds nonexchange chromosomes together at their centromeres after all other chromosomes are paired and synapsed with homologous partners.

Selected Publications