|
Gene deserts are long stretches of DNA between
genes that were once thought to have no biological function, and were
dismissed as "junk DNA." As scientists probe deeper into the DNA's
double helix, however, they are discovering that many of these "non-coding"
segments actually play an important role in regulating gene activity.
Research last year at the U.S. Department of
Energy's Joint Genome Institute (JGI) and Lawrence Berkeley National
Laboratory (LBNL), for example, has shown that gene deserts contain
DNA sequences that can switch genes on and off over considerable
distances along the DNA molecule.
Paradoxically, the same JGI and LBNL scientists
were recently able to remove giant chunks of gene desert DNA - whole
chapters in the "book of life" - from laboratory mice with no apparent
effect on the animals. Many of the deleted sequences are shared by
mice and humans, and thus may have no function in humans as well.
In an effort to resolve the paradox and to help
researchers more easily locate critical segments along the vast
stretches of DNA deserts, scientists from LLNL, LBNL, and the
Pennsylvania State University developed computational tools to
decipher gene regulation by comparing the genomes of many different
species.
When they used the tools to compare the human
genome with the recently sequenced chicken genome, they discovered
that gene deserts actually fall into two distinct categories: those
that remain relatively stable throughout eons of evolution, and those
that undergo significant variation.
Ivan Ovcharenko, a bioinformatics scientist in
LLNL's Computation Directorate who led the study, said that many lines
of evidence show that the stable desert regions, which are able to
resist genomic rearrangement and fend off infringement by repetitive
segments of junk DNA, are home to a surprisingly large fraction of the
genome's non-coding regulatory elements.
"There are many indications that stable gene
deserts represent treasure boxes of multiple gene regulatory elements,
guarding the proper complex function of the flanking genes," he said.
The variable regions, on the other hand - which
make up about two-thirds of the gene deserts and as much as 20 percent
of the entire three-billion-base-pair human genome - "can be devoid of
biological function, suggesting that a significant fraction of the
genome may not be essential.
"This information is very important for researchers
looking for mutations leading to diseases," Ovcharenko said, "because
it highlights large areas of the genome that are not likely to be
involved in causing diseases."
JGI Director Eddy Rubin, a geneticist and physician
who led the earlier studies indicating the presence of both
functioning and non-functioning gene deserts, agreed: "If you're a
gene hunter looking for genetic clues to diseases," he said, "it might
be a good idea to steer clear of regions that are devoid of critical
functions or importance," such as those that were deleted in the
JGI-LBNL mouse experiments.
The gene desert research is among a number of
studies being published today in Genome Research online and Nature
resulting from the public release of the chicken genome sequence
earlier this year. LLNL and JGI researchers, led by Lisa Stubbs in
LLNL's Biology and Biotechnology Research Program and Susan Lucas at
JGI, contributed to the Nature paper describing the primary sequence
and comparative analysis of the chicken genome, along with Ovcharenko
and Laurie Gordon of LLNL and Tijana Glavina and Andrea Aerts of JGI.
Stubbs and her team joined with JGI to compare
human and chicken genomes in 2002 and focused on human chromosome 19,
one of three chromosomes sequenced by JGI as the Department of
Energy's contribution to the Human Genome Project. The LLNL/JGI team's
high-quality sequence of parts of the chicken genome related to
chromosome 19 was used in two of the papers published today in Nature.
Participating with Ovcharenko in the gene desert
study and in the development of a new sequence comparison tool called
Mulan - shown to be instrumental in deciphering evolutionary clues by
comparing multiple primate, mammalian and fish species with chicken -
were Stubbs and Gabriela Loots of LLNL, Marcelo Nobrega of the
Genomics Division at LBNL, and Ross Hardison, Webb Miller, Belinda
Giardine, Minmei Hou and Jian Ma of the Pennsylvania State University.
The papers describing both projects will appear in the January print
issue of Genome Research. |
