A Model for Humans

This is an incredibly exciting time to be a scientist, and to be alive," Keith Cheng said. Just days before his Frontiers of Science lecture, twin papers in Science and Nature had laid out the human genome sequence in its entirety.

pink and gold micrograph with black background

"Since we know the sequence of nearly all human genes," he said, "our job has turned to the discovery of functions for all those genes. Now will be the time to use model systems—like yeast and worms and zebrafish."

Cheng uses zebrafish to learn why some mutations lead to cancer and others do not. He showed his audience a slide of a cancerous human brain: The tumor looked cloudy and ominous, obscuring most of the left hemisphere. "I try to remember this image every day to remind me why I do my work," he said, turning sober. He recently lost a close friend, his graduate adviser, to the same type of cancer.

Cancer is caused by mutations, mistakes in the sequence of one or more genes in a cell. Mutations can be hereditary or random, triggered by diet or environment, and can affect many different functions within a cell. Tumors are essentially a group of cells that grow at an uncontrolled rate. A benign tumor does not spread, and grows slowly. To become deadly, a tumor must invade the tissues around it or metastasize, spreading to another location in the body.

Numerous mechanisms within a cell keep cancer in check. Some stop cell growth, while others act as spell checkers, correcting small mutations in genes. A major factor in nearly all human cancers is the loss of function of tumor suppressor genes. Tumor suppressors can be disabled by mutations in other genes that control the stability of a cell.

Cheng and postdoctoral scholar Jessica Moore wanted to better understand this process and its effect on cancer susceptibility. But they needed a model.

green and brown and gold micrograph

Enter zebrafish. Why use these miniature striped creatures for cancer research? They are less expensive to raise than rats or mice, and are easy to keep. They are vertebrates, and can get cancer, like humans. They generate hundreds of eggs per breeding pair every week, giving numerous chances to test and retest hypotheses. Best of all, the embryos, unlike ours, develop separately from their mothers, and are completely transparent. Individual cells can be easily observed under a microscope. As model systems, zebrafish provide a clear view to what might be happening in humans.

Cheng's approach is called a "mutant screen," an example of forward, or classical genetics. He damages random genes and then identifies which characteristics of the fish are different after the mutations. The opposite approach, reverse genetics, works by guessing that a gene has a specific function, and then knocking out that gene to observe what happens. In his lecture, Cheng compared forward genetics to a Martian who wants to find out what makes cars stop. The Martian randomly damages thousands of cars and screens for cars that no longer stop. The Martian finds that brake pedals, hydraulic lines, master cylinders, and brake pads all contribute to making a car stop. Like the Martian, Cheng randomly damages genes in zebrafish and studies the mutants. He looks for "genomic instability" mutations—mistakes that yield even more mistakes in other genes. When he discovers which genes were damaged and which abnormalities resulted from each type of damage, he will better understand how genetic damage occurs, and perhaps learn how to prevent it.

Using this methodology, Cheng sought to generate fish whose eyes are sprinkled with white spots, a phenotype called "mosaic eyes." These fish are known to have "unstable" genomes, and model the genetic events that occur in human cancer. The process begins with a normal female fish and a male that has been exposed to a chemical mutagen. Their offspring receive normal genes from the mother and mutated genes from the father. The second-generation female eggs are then fertilized by sperm that have been irradiated by ultraviolet light, which effectively kills the DNA. Offspring of this "parthenogenic" mating are born with only one set of functioning DNA—that of the mother—which then allows Cheng to identify mutations carried by the mother. Fish with "mosaic eyes," for instance, have eye cells in which the pigment genes were mutated. By using parthenogenesis, Cheng knew the mothers of the fish with mosaic eyes had unstable genomes.

Cheng and his team's effort to discover their first mosaic eye mutant lasted five years. "The day we found our first mutant was one of the most exciting days of my scientific life," said Cheng.

Some months later, tumors began to appear in fish carrying the new mutations. Cheng and Moore were thrilled because the appearance of tumors suggested that their hypothesis—that mutations in certain genes that control genomic stability cause susceptibility to tumors—was correct. Though exciting, Cheng warned that they must study larger numbers of fish to prove that specific gene mutations directly cause tumor susceptibility.

Since then, Cheng has developed a method to speed up the screening of thousands of embryos for possible mutations. It involves a mold created with surgery resident Gladys Tsau-Wu, in collaboration with a machinist in Penn State's artificial heart program. The mold enables Cheng to study 64 zebrafish embryos at once, sectioned on small glass slides, the same way a pathologist studies cancer tissues. Through these screens, Cheng, Moore, and Tsao-Wu have found mutants with names like "tumor head" and "huli hutu" (which means "confused" in Mandarin Chinese). The names might be humorous, but no one forgets the importance of the fish. As models for humans, zebrafish offer great insight into human cancer.

Keith Cheng, M.D., Ph.D., is associate professor of pathology and a member of the Jake Gittlen Cancer Research Institute in the College of Medicine, Milton S. Hershey Medical Center, 500 University Dr., Box 850, Hershey, PA 17033; 717-531-5635; kcheng@psu.edu. Jessica Moore, Ph.D., is a postdoctoral scholar at the Jake Gittlen Cancer Research Institute; jmoore@psu.edu

SIDEBAR

The Next Step

Last Updated September 01, 2001