Science Editor

The mom or dad (Homo sapiens) who goes into a pet shop to buy the kids a couple of minnow-sized, brightly striped zebrafish (Danio rerio) is separated from his or her finny purchase by 400 million years of evolution. Mice (mus musculus) are four times closer to human homology; those two species diverged only 100 million years ago.

Mice, of course, are the staple animal model of biological science. To be sure, fruit flies (Drosophila melanogaster) and nematodes (Caenorhabditis elegans) play big roles in in vivo research, but unlike mice and zebrafish, they lack backbones. It takes a vertebrate to match genes with the top vertebrate - humans.

"There are a lot of genes in the human genome - 30,000 at least," observed molecular geneticist Nancy Hopkins, at the Massachusetts Institute of Technology in Cambridge, Mass. "But the number that play critical roles in any biological process is small. To discover what those genes actually do, you've got two choices: Go either to forward genetics or backward genetics.

"In the usual backward genetic approach," she went on, "you would take all the genes in the genome, knock them out one at a time, and see what happens. The other approach," Hopkins continued, "is what we do, which is to go forward. We mutagenize the genome blindly, then select mutations that have something wrong with the process we're interested in. To do that with the zebrafish genome, we picked a handful of genes out of who knows how many genes a zebrafish has - probably, like a human, 30,000 to 60,000.

"But we've found and cloned just a small number of the fish's total gene complement - 75 so far - required to create a baby zebrafish," Hopkins added. "It's estimated at 2,400 found that affect the development of its first five days of life. Of all the genes in the zebrafish genome, the ones that do the major job are quite few. There's somewhere between 100 and 150 genes just to form each organ. But that's a tiny number out of 30,000 to 60,000 in the genome. The most efficient way to hunt for those first 75 genes," she recounted, "was putting our hands on the ones that actually formed those organs in the vertebrate organism."

Recipe For Insertional Mutagenesis

Hopkins is senior author of a paper in the May 2002 Nature Genetics, released online May 13th. Its title: "Insertional mutagenesis in zebrafish rapidly identifies genes essential for early vertebrate development."

"For forward projects," she explained, "you have to do mutations. You can use chemicals, which is the more common way. However, that method creates single-base chains in the DNA, so it's hard to find the genes. We used an insertional mutagen, which is a little piece of DNA. In this case it's a viral vector that comes in from outside. It's the Moloney mouse leukemia retrovirus - which can't replicate - wrapped in a vesicular stomatotis virus subunit. This construct randomly integrates its own DNA into the middle of the zebrafish gene. In doing so, it disables that gene by mutating it.

"And it stays there," she continued, "serving as a permanent little molecular tag on the gene that's been damaged. Some call it gene tagging; we call it insertional mutagenesis when we cause a mutation that way," Hopkins pointed out. "We were able to use the exogenous marked sequence as that tag, to pull it out of that region and its surrounding DNA locus very rapidly - in as little as two weeks. That made it more straightforward for identifying the mutated gene than using a chemical mutagen - the standard way - which lacks a tag."

In the course of finding and analyzing the zebrafish genes that orchestrate its organ development, Hopkins and her co-authors encountered counterparts of human disease genes.

"We had a paper earlier this year from a postdoc in my lab named Zhaoxia Sun," she recalled. "And her paper, I thought, was pretty cool on this exact topic. She's interested in the kidney, and studied a zebrafish kidney mutant. When the gene was cloned it turned out to be a transcription factor - already thought to be the cause of an inherited form of human diabetes. In that particular genetic disease, people with diabetes also have kidney defects. Then Sun looked at the pancreas to see if there was something wrong there. Indeed, the pancreas was not developed properly, so insulin could not be expressed properly in the fish.

"Each of the 75 genes had some similarity with a known human gene," Hopkins observed. "Eventually, our team will have identified 450 to 500 genes. So the disease information in the human told us what to do next in the zebrafish. The similarities are remarkable. But the fish is not a human, so hopefully there are differences, too. But in terms of human disease I think there are a number of examples. We have people studying mutations affecting hearing in zebrafish. And those genes turn out to be ones that have been identified in human and mouse disorders of hearing. Certainly in many cases there will be similarities," Hopkins observed. "How else could you identify the genes that are essential for cell survival and growth? I think they could be potentially powerful as a set of possible targets for treating cancer or other human diseases."

Do Early Organ Genes Work In Adulthood?

"And then, of course, in terms of identifying the genes needed for growing embryonic organs," Hopkins continued, "the question is whether those same genes are needed when you replenish organs later in adult life. This set of genes are expressed heavily in the embryo, but not in a whole adult. Can these embryonic genes get reactivated when you replenish cells later on? It's not unlikely. The genes that really stop things dead, or do something dramatic in a given process, are the ones likely to have some function.

"Patent applications have been filed on some of these novel genes for making organs," Hopkins said. "More work is needed to make them really valuable. One of my former postdocs, Wenbiao Chen, and his partner, Roger Cohen, have just started a small biotech company in Portland, Ore., based on this technology. It's named Znomics Inc."