MIT model explains how the brain can learn novel tasks while still remembering what it has already learned.
CAMBRIDGE, Mass. -- By inducing fatal mutations in zebrafish embryos, a researcher at the Massachusetts Institute of Technology is determining which of the fish's estimated 100,000 genes are truly indispensable, thereby pairing specific genes with the growth of specific body parts in a way never before possible.
This work by Nancy Hopkins, Amgen professor of biology, has been called unique. It could lead to new drugs and therapies for humans, many of whose genes are similar in sequence to those of zebrafish. Humans, zebrafish and other vertebrates share many identical or near-identical genes.
Identifying an embryo's crucial genes by introducing mutations and tracking the outcomes of the mutations in future generations is a project of epic proportions. The initial report of Hopkins's findings will appear in the October 15 issue of the journal Genes and Development.
If Hopkins can identify which genes are crucial for normal development, she will have added enormous value to the Human Genome Project, which, when completed, will have identified and mapped the roughly 100,000 human genes. The genome project, while identifying the sequence of all 100,000 genes, cannot tell researchers what the genes do or which set of genes, for example, work together to develop a normal heart or a healthy nervous system.
Most genetic research in vertebrates is reverse-engineered, meaning that scientists start with a gene and work backwards, mutating the gene to determine the effect of removing just that one gene. Forward genetics, or planting a mutation at random and seeing how it plays out in future generations, however, is the most powerful approach to discovering gene function, Hopkins said. "This way, you let biology tell you what the genes are doing rather than trying to be smarter than nature."
Hopkins and her team have isolated 45 mutations so far. She hopes to identify 1,000. Among the 45 already in hand is one that results in fish that fail to grow to normal size. They are tiny, perfect replicas of normal adults. A mutation in another gene causes fish to have fins that don't stop growing. Yet another mutation causes fish embryos to curl up because their spines don't form properly. Some mutants aren't the right color. Some have defective livers. Others have muscles that form normally but then appear to fall apart.
While not all these discoveries will result in new therapies, some are expected to. California-based Amgen, which supports Hopkins's research, recognizes its potential. Amgen would like to get first crack at Hopkins's results in case she finds, for instance, a gene that specifies the production of a medically valuable protein.
Someday, it may be possible to replace a damaged gene in utero, curing a potentially fatal condition before a baby is even born. Or introduce a gene that could prevent or cure cancer. Yet other applications may involve growing new tissues that replace those that wear out.
But you have to know which gene is the right one for the job.
A MODEL SUBJECT
There were two major hurdles to doing this kind of experiment, which an MIT Nobel laureate describes as the only one of its kind. One is the problem of inserting a genetic mutation into the zebrafish that can be traced to a specific gene. The other is the logistical nightmare of raising so many generations of fish until the defects are apparent. Defects are seen in fish that are the are the great-grandchildren of those whose genes were manipulated.
In 1996, Hopkins, with a postdoctoral fellow and a graduate student, found a mouse leukemia virus that would infect zebrafish embryos and leave them with easily identifiable, marked genes. Others attempting the same process have used a chemical bath, which also damages genes, but in a way that, in contrast to Hopkins's method, leaves each damaged gene extremely difficult to identify.
Unlike experiments that involve plants or insects where forward genetics has been highly successful in the past, Hopkins' experiment is particularly valuable because it involves vertebrates. Zebrafish are used for this kind of experiment because in only a few hours, their eggs develop into visible, transparent clusters of cells that are easy to work with. Researchers can almost tell with scarcely more than a magnifying glass whether a genetic mutation has occurred.
THREE GENERATIONS OF FISH
When you see the rows upon rows of fish tanks in Hopkins's laboratory filled with dozens of fish ranging from the tiniest larvae to full-grown adults, it makes Gregor Mendel's peas seem like child's play. Hopkins estimates that one and a half million fish will pass through her lab during the three-year project.
No scientist would ever downplay the importance of Mendel's discovery. But few have undertaken a genetics project of this scope indoors and in a university setting. "People, including ourselves, questioned whether we could ever actually do it," Hopkins said.
The reason the project is so huge stems from the way genetic information is passed from one generation to the next.
As Mendel discovered, a recessive trait, for example cystic fibrosis in humans, when inherited from one parent is carried in the genetic material of one-half of the offspring. Those offspring that carry the mutation, when mated with siblings who are also carriers of the same genetic defect, will have offspring expressing the trait. In the Hopkins lab scheme, it takes three generations for a genetic mutation to show up.
To generate 35,000 mutated fish to use in the experiment, Hopkins and her colleagues performed insertional mutagenesis on a quarter of a million fish eggs. They believe that together, the fish harbor nearly 1 million insertions in their genes "Generating that number of lesions in a vertebrate genome is almost unimaginable," she said.
The genetically manipulated fish in Hopkins's lab are no ordinary fish. They are probably some of best cared-for fish in the world. If one dies, the troops are called in to nose out the problem. A fish consultant visits them weekly to check the water quality and filtration of their 4,000 connected tanks.
The tiny larvae are fed, like human babies, all day long, seven days a week. Because they must all grow under the exact same conditions, each baby fish is fed the exact same number of paramecia. Hopkins, who admits that at one time she had trouble keeping pet store fish alive, has a staff of 25 to care for and help analyze the genes of the fish to determine which ones contain the mutations and which ones are healthy. "We call this 'The Fish Ritz,'" Hopkins said.
It's a lot of work, but Hopkins is sure she is onto something important.
Her colleagues agree. This project "is unique. No one in the world is in a position to do what Nancy Hopkins is doing," said Phillip Sharp, Institute Professor of Biology and MIT and Nobel laureate. "That this rapidly reproducing vertebrate system can be analyzed by identifying genetic material for all the processes in vertebrate development is really exciting. There's nothing like it."
Hopkins's team includes postdoctoral associates Adam Amsterdam, Shawn Burgess, Wenbiao Chen, Greg Golling and Zhaoxia Sun; technical associate Sarah Farrington; research scientist Maryann Haldi; postdoctoral fellows Ernesto Maldonado-Olvera and Marcelo Antonelli; and MIT affiliate Karen Townsend. This work is funded in part by Amgen.