MSKCC's Keeney Describes DNA-Strand Gymnastics

The summer Olympics may be a distant memory, but strands of DNA continue to perform their own brand of gymnastics. During meiosis, where gametes like sperm and eggs or yeast spores are made, strands of DNA from separate chromosomes cross over each other and sometimes exchange places. Dr. Scott Keeney of Memorial Sloan-Kettering Cancer Center in New York came to NIH recently to talk about the different kinds of strand gymnastics that determine whether those pieces of DNA will ultimately change places.

Dr. Scott Keeney of Memorial Sloan-Kettering Cancer Center in New York came to NIH recently to talk about different kinds of strand gymnastics.

The technical name for this exchange of DNA is homologous recombination. The cell uses it to repair DNA that's been damaged, and to help make sure that pairs of chromosomes separate properly during the first division in meiosis. Homologous recombination also gives resulting gametes more genetic variation. When the process goes awry, it can result in cancer and fertility problems. Scientists also think it's one of the leading causes of developmental disabilities.

Keeney called the process a "highly regulated pathway of self-inflicted DNA damage." He explained how both strands of one DNA molecule are broken to make double-strand breaks (DSBs), and how these breaks are then repaired.

When he started in this field, Keeney said, researchers already knew all the players that were involved in making the DSBs. But now, 11 years later, the protein called Spo11 is the only one whose function has really been pinned down biochemically. Keeney's group identified Spo11 as the one responsible for cutting the DNA strands to begin the process. Spo11-like proteins are found in many other species, from yeast to mouse to human. For these studies, Keeney's laboratory used the single-celled "budding" yeast, commonly used to make bread and beer, because of the ease with which it can be genetically manipulated.

Other studies have identified at least 9 other proteins working with Spo11 to cause DSBs, but the functions of these proteins are not known. Therefore, Keeney's group set out to explore systematically the interactions between Spo11 and these other proteins in order to gain insight into their roles.

"One of the take-home lessons," Keeney said, "is that there are connections between all these players."

A protein called Ski8 emerged as the one with the strongest interaction with Spo11. Ski stands for Super Killer because the experiments that originally identified the protein, which were performed by Dr. Reed Wickner's group at NIH, resulted in a lethal number of RNA viruses proliferating in cells. While Ski8 plays a role in RNA metabolism out in the cytoplasm, it also works with Spo11 in the nucleus during meiosis to cause DSBs. Keeney's group found that a direct interaction is required between the two proteins for DSBs to form. Ski8 seems to work with Spo11 in recruiting other DSB proteins to the chromosomes during meiosis. Keeney spoke about some of these other proteins, but their exact roles in homologous recombination are still unknown.

Keeney, however, is most interested in discovering how the cell makes the decision whether or not to exchange the DNA strands when repairing a DSB. Many more DSBs are made than crossovers completed, he said. After a DSB is made, the cell must somehow decide whether to recombine the strands or not — in other words, which kind of strand gymnastics to perform.

There are essentially two types of models explaining how the cell might do this, Keeney explained: counting models and physical models. In counting models, the cell somehow counts a specific number of non-crossovers for each crossover. In a physical model, a crossover would somehow cause a signal to spread along the chromosome, preventing other crossovers.

Keeney's team figured out a way to test these models by using a series of Spo11 mutations that altered the level of total DSBs. If the counting model is right, the number of final crossovers would vary with the number of DSBs. If a physical model is right, the number of final crossovers wouldn't change much over a range of DSB frequencies.

Keeney's team found that as DSBs were reduced, the crossover frequency showed no consistent change. This clearly contradicts the counting models; however, what the physical signal might be that causes this effect is still open to speculation. Whatever the mechanism, Keeney believes that the number of crossovers in a cell needs to be preserved to ensure cell survival — a process he calls crossover homeostasis. When the number of DSBs drops below a certain level, yeast spores simply can't survive.

"What does the cell care about?" Keeney asked. "I don't think the cell really cares about the relative distribution of crossovers and non-crossovers." Getting the right number of crossovers, he argued, is what's most important. Once that's done, the cell can worry about dealing with the rest of the DSBs.