Some 400 'Fragile Regions' Of Genome More Vulnerable To Evolutionary Breaks; New Theory Could Have Implications For Cancer Research

"These rearrangements are like earthquakes that are more likely to happen along fault lines, which is why you're more likely to see a quake in Los Angeles than Chicago," said Pavel Pevzner, the Ronald R. Taylor Chair in the Jacobs School's Computer Science and Engineering department, who co-authored the study with project scientist Glenn Tesler. "Similarly, there are 'faults' within the human genome. They are fragile regions, as opposed to solid regions that show much less propensity for rearrangement and make up about 95 percent of the genome."

Pevzner and Tesler are experts in bioinformatics -- the use of computing and mathematics to study genomes -- and their study grew out of the first giant genomic sequencing projects, funded in part by the National Institutes of Health. Those projects resulted in DNA sequencing of the human and mouse genomes. In December 2002, that comparison led the scientists to compute 281 large blocks (of one million 'letters' or more, out of roughly three billion letters in the mammalian genome), and 245 major rearrangements since the two species evolved from a common ancestor 75 million years ago. "Like the ancient super-continent Pangea broke into seven continents 130 million years ago, the genome of the common mouse and human ancestor that lived 75 million years ago broke into 281 blocks in the course of human-mouse evolution," added Pevzner. "It's like starting out with two decks of cards representing the ancestral genome, and the decks are re-shuffled so the cards end up in a different sequence."

Biologists have assumed that the re-shuffling is random -- a view enshrined in 'random breakage theory,' which has been borne out by many major studies over the past thirty years. "Based on random breakage theory, the breaks in the genome should have been evenly distributed," said Tesler. "Instead, they happened surprisingly often in these fragile regions, and avoided other regions altogether."

"Now that we are able to compare the genomes, it becomes clear that these breaks are not random," concluded Pevzner. "They occur disproportionately in these fragile regions, just as earthquakes happen more often near major fault lines."

Pevzner and Tesler do not discount random breakage theory. In fact, pointed out Pevzner, that theory "was prophetically accurate in predicting roughly 200 major genomic blocks in human-mouse evolution -- a prediction made in the mid-1980s, years before sequencing of the two species' genomes was complete." At the macro level, their new theory leads to the same conclusions as random breakage theory, and thus is consistent with all genetic data observable prior to 2002. However, at the microscopic level, when it comes to the short breakpoint regions, their theory explains the new genomic data while random breakage theory fails to do so.

The UCSD scientists hope to gather more evidence for their 'fragile breakage theory' by running similar genomic comparisons with new genomes as they are sequenced -- first rat, and later cat, dog and other mammals. They also want to analyze genomic data for clues to understanding whether the major rearrangements happen all at once, or are the product of multiple smaller rearrangements over time. "It's like looking at earthquake damage in the San Fernando Valley," observed Pevzner. "Was it the product of the 1994 Northridge quake alone, or the result of many smaller quakes over time?"

Meanwhile, Pevzner and postdoctoral researcher Ben Raphael are collaborating with biologists at the University of California, San Francisco (UCSF) Cancer Center to do computational analysis of genome rearrangements implicated in breast cancer. "Surprisingly enough, you can view the breast cancer genome as an extremely fast-evolving human genome," said Pevzner. "All genomes have fragile regions that are more susceptible to rearrangements. Obviously the time scale is very different, as are the consequences -- disease and even death from cancer, or the birth of new species or improvements to existing species from evolution. But we are not yet sure whether cancer uses the same type of fragile region to break DNA as used in human evolution. We are working on it and we already have the first rough draft of the genomic architecture of certain breast-cancer cell lines."

Pevzner and Tesler also hope to drill down deeper into the genomic data, to analyze how their fragile breakage theory works when looking at smaller blocks of genetic letters. "It's not like looking at a magazine, where the layout tells you exactly where one story breaks and another begins," explained Tesler. "With the human genome, we know where chromosomes break, but we don't know how big all its component units are. So the best we can do is look at large quantities of data and use computational methods to discern patterns."

Concluded Pevzner: "I am confident that within a few years, we will be able to figure out where the 'faults' are in the human genome, and that will give us much greater insight into the nature of evolution, and hopefully a better understanding of cancer and other diseases."