A science-fair judge turned Constance Cepko into a biologist. The courage to switch fields led her from virology to retinal development. And the realization that basic biologists can make inroads into medicine inspired her to study a common eye disease, retinitis pigmentosa, in which the retina's light-sensing cells degenerate. "It turns out that we find out how disease genes work in humans by looking at how they work in model organisms," she explains.
When Cepko was in seventh grade in Laurel, Maryland, she had to submit a science-fair project. Taking cues from a children's magazine, she experimented with baker's yeast. The venture was so successful that the judge, John Palmer, invited her to work in his research lab on Saturday mornings, which she did for the next four years. "It's because of John that I'm where I'm at now," she says.
Cepko moved from yeast to viruses to obtain her Ph.D. at MIT. Then as a postdoc at MIT, she made retroviral vectors, which are used to transfer genes in basic science or disease studies. But when she established her own lab at Harvard Medical School in 1985, she wanted to use those tools to address big, unanswered questions. These seemed to abound in neuroscience, her reading revealed. "And I thought the retina would be an incredibly interesting tissue," she says. "So I just picked it."
The retina, a protuberance of the brain, consists of several thin layers of tissue. Like a camera, this tissue records the direction, intensity, and frequency of light waves. Like a computer, it processes that information and sends it through the optic nerve to other parts of the brain. As the retina develops, its progenitor cells (the offspring of embryonic stem cells) have to decide when to divide, what types of nerve cells to become, and how many of each type to form. The nerve cells then have to interact with each other to form the computer circuits. These processes occur in all nerve tissue, but the more accessible retina can be studied more easily than other parts of the brain.
When Cepko began to investigate retinal development, it was assumed that different types of neural progenitors gave rise to the various types of neural cells. But by labeling cells with markers that persisted from one generation to another, Cepko's first graduate student discovered that neurons and housekeeping cells called glia come from the same progenitor cells.
Progenitors might make different daughter cells because they become different as development progresses or because the environment changes over time. Studies of progenitor cells dividing early and later in development showed that both scenarios are correct. "Early and late progenitors do respond differently to environmental cues, and environmental cues can influence the kind of fate choices that are made," Cepko says.
The group discovered that new environmental cues do not trigger the production of different types of cells. Instead, early progenitors alter the ratio of cell types that are needed for early development, and late progenitors alter the ratio of cell types appropriate for the late stage. Cepko's group is currently studying the molecular basis for these intrinsic differences between early and late progenitors. "But it's going to be a very long and complicated story," Cepko says.
In the mid-1990s, Cepko decided that it was unrealistic to study cell fate without considering the influence of patterning genes, which set up the basic coordinates of the retina, such as the anterior/posterior or dorsal/ventral axes. "Eventually, a cell finds itself in the crosshairs of the patterning system," Cepko says. "So when it decides what to be, part of its decision is based on where it sits in the tissue."
The human retina has a very pronounced pattern: a central area called the macula that is rich in cones (photoreceptor cells that function in bright light and allow us to see color) and a peripheral region rich in rods (which operate in dim light). Looking for an experimental system, Cepko discovered that chickens have a similar stark arrangement of rods and cones. By using retroviral vectors, she has now identified a number of chicken genes that alter retinal patterning if they are expressed at abnormally high or low levels. The big question is, "How does the retina know to put the rod-free zone smack in the middle?" Cepko asks. She says it will be necessary to understand the molecular cues that tell cells to become cones or rods before this question can be answered.
Until a decade ago, Cepko thought she had little to contribute to medical science. But basic studies with mice, and even with worms and flies, revealed that many genes have been conserved throughout evolution. "The realization that work with model organisms has a direct applicability to human disease has enabled me and other scientists to recognize that the things we know and the techniques we have developed can be brought to bear more directly on some of these problems," Cepko says. She is now identifying genes that, when faulty, contribute to retinitis pigmentosa or macular degeneration, the leading cause of blindness in older people. For example, her group is studying diseases in which intrinsic defects in a rod gene cause rod cells to die. By understanding why the death of cones follows that of the rods even though the cones' genes are not defective, they hope to generate new ideas for treating macular degeneration. "The methods of cellular and molecular biology have grown so powerful that we can now try to understand the mechanisms of human disease more directly," Cepko says. "So that is what we are doing."