WITH SEVERAL GENOME SEQUENCING PROJECTS COMPLETE, scientists now know where to find all the genes for organisms ranging from humans to fruit flies to yeast. But knowing where a gene is found and what information it typically contains still leaves two critical questions: What jobs does the protein made by the gene do? Where in the organism does it do those jobs?

As a result, geneticists interested in studying genes linked to a specific structure or disorder — the heart, for example, or heart disease — sometimes face months or years of grinding, repetitive search and analysis to identify the specific genes they want to study.

Susan K. Dutcher, PhD, professor of genetics and of cell biology and physiology, recently developed an innovative way to quickly pick out genes of interest. Her approach combines modern computing power and millions of years of evolutionary history built into DNA to sift through thousands of genes and selectively extract many genes related to her research.

The new approach has been spectacularly successful, quickly leading to benefits for both Dutcher’s research program and for other researchers, including a group at Johns Hopkins University that used her data to identify a human disease gene.

Susan K. Dutcher, PhD, and Gary D. Stormo, PhD, with research assistant Lin Ya Li (left, seated) and graduate student Jin Billy Li.

Dutcher’s new method centers on a computerized comparison of the genomes of three species: humans; a weed, Arabidopsis; and Chlamydomonas, a green alga.

Dutcher works with Chlamydomonas to learn more about cilia, hair-like structures on the surfaces of cells.

“Almost every cell in the human body has cilia,” Dutcher says. “Cilia that are active early in development ensure that organs like the heart and stomach end up where they’re supposed to be. Cilia clear away dirt and bacteria in the respiratory tract, help sperm swim and help keep fluid flowing into and out of the brain.”

Problems in the cilia and basal bodies, the structures that anchor cilia to the surfaces of cells, are linked to a variety of disorders, including polycystic kidney disease and genetic disorders that affect the ear, nose, sperm, placement of internal organs, number of fingers and toes, and length of the limb bones.

Studying Chlamydomonas allows Dutcher’s group to more easily isolate and manipulate cilia and basal bodies. Although the alga is more easily subject to experiments, its genetic material has 20,000 genes, which still left researchers like Dutcher with quite a bit of searching to do. Amid those thousands of genes, Dutcher and other scientists were seeking only an estimated 250 to 400 genes — potentially a very long and repetitive hunt.

“It had been pretty slow going,” Dutcher says. “Although there are a lot of people interested in Chlamydomonas and in these genes, we had only identified a few of them at the time we began this genome comparison.”

Dutcher’s analysis derives its power to identify interesting genes from the evolutionary history written into the organisms’ genomes. Evolution tends to retain genes that make proteins for essential cell structures and processes. Scientists call genes kept intact through the development of many different species “highly conserved,” and Dutcher and others have shown that genes for cilia and basal bodies fit this description.

Significant exceptions to this principal of conservation can occur when life has to make major adjustments to its environment. In one such instance, most plants evolving to adapt from life in the sea to life on land discarded their cilia.

“If you look at any flowering land plant, it doesn’t make any cilia, and it doesn’t make basal bodies,” Dutcher explains. “Once the idea hit me, it seemed fairly obvious: Why not have the computer look for gene matches between humans and Chlamydomonas, and then compare those results to a land plant and remove all the genes from the first comparison that also had matches in the land plant?”

The human-to-Chlamydomonas comparison would locate many of the genes for basic functions and structures those two organisms share, Dutcher reasoned, but any matching genes also found in the flowering land plant likely would have nothing to do with cilia or basal bodies.

“We thought the final result might be incredibly enriched with cilia and basal body genes,” she explains.

Dutcher approached Gary D. Stormo, PhD, professor of genetics, with her idea. “My first reaction was, that’s an interesting idea, but that’s a lot of genes!” Stormo recalls.

A specialist in computational analyses of genetic code, Stormo is particularly interested in finding signals contained in DNA that interact with proteins to turn genes on and off.

Stormo and Dutcher, who are married, were jointly mentoring a graduate student, Jin Billy Li, who took the lead in carrying out the comparisons. For the land plant, researchers chose to use the genome of Arabidopsis, a weed. (Scientists at Washington University’s Genome Sequencing Center and at Cold Spring Harbor Laboratory completed mapping of the Arabidopsis genome in 2002.)

The human-to-alga comparison produced 4,348 gene matches. Comparing those results to Arabidopsis yielded approximately 3,600 matches, allowing researchers to narrow down the pool of prospective cilia and basal body genes to 688. Further comparison with the genomes of the fruit fly and the sea squirt, a small ocean-going animal, reduced the list of genes to approximately 300.

To determine how effective the screening process had been, Dutcher and colleagues first looked to see if their results had highlighted any of the cilia and basal body genes scientists had already identified in Chlamydomonas. They were delighted to find the comparison caught 90 percent of the genes they already knew about. “This absolutely flabbergasted us,” says Dutcher. “We thought we’d be less successful.”

Other tests produced similarly encouraging results. Li and research assistant Lin Ya Li cut the cilia off Chlamydomonas and watched to see if 103 of the genes identified by the comparison became more active as the algae rebuilt the cilia. More than a third increased their activity. A follow-up experiment suggested that several genes that didn’t increase their activity when the cilia were cut off were instead involved in the construction of basal bodies.

“We don’t yet know how many of the genes we identified are completely unrelated to cilia and basal bodies, but so far the results have been very encouraging,” says Stormo. “The technique seems to have been not just a moderately good filter, but in fact a very good filter for the genes Susan is interested in.”

Dutcher then searched for matches between the genes found by the comparison and known human genes for cilia and basal bodies. She found that the research team’s results had highlighted both of the genes associated with juvenile polycystic kidney disease and five of six genes linked to Bardet-Biedl Syndrome (BBS), a rare genetic condition that causes blindness, mental retardation, severe obesity and other problems.

Dutcher became intrigued by the possibility that the comparison’s results might help scientists identify new disease genes. She learned that scientists had narrowed the hunt for a seventh BBS gene down to a large region on human chromosome 2. However, that region had approximately 230 genes, far too many for scientists to examine on a one-by-one basis.

“We contacted Nicholas Katsanis, PhD, a BBS researcher at Johns Hopkins, and he told us they had no idea where the BBS gene on chromosome 2 was,” Dutcher recalls. “We explained that our comparison
had highlighted two genes in that area and suggested they might be worth checking into.”

When Katsanis, assistant professor at Hopkins’ McKusick-Nathans Institute of Genetic Medicine, analyzed the two genes in families afflicted by BBS, he found several sufferers had abnormalities in one of the genes and named it BBS5.

To confirm that the BBS5 gene was important to cilia or basal bodies, Dutcher reduced the gene’s activity in the alga.

“We see different effects depending upon how much we knock the activity of the gene down, but it looks like it actually is a basal body protein,” Dutcher says. “Our hope is that we can now take our data set and accelerate the search for genes that contribute to some of the many human disorders that involve cilia and basal bodies.”

In addition to probing those potential connections, Dutcher already has started brainstorming new genetic comparisons designed to identify genes of interest.

“Humans have two kinds of cilia — motile cilia, which create motion, and non-motile cilia, which respond to motion,” she says. “The microscopic worm C. elegans only has non-motile cilia, so if we were to take our results from this study and eliminate all the genes that have a match in the genetic code of C. elegans, that might let us highlight genes for proteins that create and control the movements of cilia.”

Mark Johnston, PhD, professor and head of the department of genetics, says Dutcher’s results have other geneticists excited about the possibility of using similar comparisons to identify genes that carry out important functions or contribute to human diseases.

“What’s so heartwarming about this result is how vividly it illustrates the value of basic genetic research in these model organisms,” says Johnston. “There are quite a few scientists now thinking about how we can use this approach in other contexts.”