Mysterious snippets of DNA withstand eons of evolution

- By Krista Conger

Gill Bejerano

Small stretches of seemingly useless DNA harbor a big secret, say researchers at the School of Medicine. There's one problem: We don't know what it is. Although individual laboratory animals appear to live happily when these genetic ciphers are deleted, these snippets have been highly conserved throughout evolution.

'The true function of these regions remains a mystery, but it's clear that the genome really does need and use them,' said Gill Bejerano, PhD, assistant professor of developmental biology and of computer science. In fact, these so-called 'ultraconserved' regions are about 300 times less likely than other regions of the genome to be lost during mammalian evolution, according to research from Bejerano and graduate student Cory McLean published in the Oct. 2 issue of Genome Research.

Although some of the ultraconserved regions, which were first identified by Bejerano in 2004, are involved in the regulation of the expression of neighboring genes, previous research has shown that mice missing each of four regions seem perfectly normal.

'It's very surprising that none of the four has any observable phenotype,' said Bejerano. 'In some ways it just doesn't make sense.'

This lack of effect is usually taken as a strong argument against an important functional role for the missing segments of DNA - either because they don't do much or because other bits of DNA serve as understudies when the primary actors are missing. But in this most recent study, evolution roars over the squeak of the seemingly contented mice.

'When we tried to determine whether similar deletions occur in the wild,' said Bejerano, 'we found that this is almost never seen in nature.'

McLean and Bejerano compared the likelihood that ultraconserved elements of at least 100 base pairs shared by humans, macaques and dogs would have been deleted in rats and mice, with the likelihood of a similar pattern in non-conserved DNA. Less than one-tenth of 1 percent of segments completely identical among the primates and dog were missing in the rodents. In contrast, about 25 percent of non-conserved segments were absent in the mice and rats.

It's not that these regions are somehow protected against change: they are mutated in about one in 200 healthy humans. Rather, these changes seem to be swept away over time by the tides of evolution in a process called 'purifying selection.' Bejerano and McLean believe that something similar may be happening in the laboratory mice on a scale too subtle to be seen under carefully controlled experimental conditions.

After establishing how infrequently the ultraconserved segments are deleted, the researchers investigated whether the degree of homology (the percent of nucleotides shared between species) or the extent of conservation (the evolutionary distance between species that share a version of the sequence) correlated most closely with the likelihood that it would be lost in primates or rodents. Sequences shared among many distantly related species are likely to be older than sequences found only in closely related species. The researchers found that less-highly conserved sequences shared among several distantly related species - including opossum, platypus, chicken, frog and fish - are more likely to also occur in humans than are more-homologous sequences that occur in only a few closely related species. The likelihood that a sequence will be found in humans increases as the evolutionary age of the sequence increases.

'Interestingly,' said Bejerano, 'the longer the sequence has been in us, the less likely it is to be lost. It's almost like the bricks in the foundation of a building, which hold up the rest of the structure.'

Clearly there remains a lot to be discovered. The upcoming availability of several additional well-sequenced mammalian genomes will give the researchers even more data with which to work. And subjecting the laboratory mice missing the ultraconserved regions to a variety of conditions, such as changes in diet or living conditions, may make more noticeable any differences between them and the mice without changes.

'Evolution is a lot of fun,' said Bejerano, who plans to continue the investigation into what the ultraconserved segments might be doing. 'You answer one question, and five others pop up. But one of the most rewarding things to me is the fact that we're developing a growing appreciation for how much these regions actually matter.'

The research was supported by a Stanford Bio-X graduate fellowship to McLean and an Edward Mallinckrodt, Jr. Foundation junior faculty grant. Bejerano is a Sloan research fellow and a Searle scholar.

About Stanford Medicine

Stanford Medicine is an integrated academic health system comprising the Stanford School of Medicine and adult and pediatric health care delivery systems. Together, they harness the full potential of biomedicine through collaborative research, education and clinical care for patients. For more information, please visit

2023 ISSUE 3

Exploring ways AI is applied to health care