Kiara Eldred sometimes compares her nine-month-long scientific experiments, growing tiny human retinas in a laboratory dish, to raising children.
Eldred, a graduate student at Johns Hopkins University, starts by growing thousands of stem cells and feeding them nutrients and chemicals that will steer them to develop into the retina, the part of the eye that translates light into the signals that lead to vision. After two weeks of painstaking cultivation, those cells typically generate 20 to 60 tiny balls of cells, called retinal organoids. As they mature, these nascent retinas get dirty and slough off lots of cells, so they also need to be washed off when they're fed every other day — at least for the first month and a half.
After nine months of assiduous care, Eldred has a batch of miniature human retinas that respond to light, are about two millimeters in diameter and are shaped like a tennis ball cut in half. But growing the organoids is only the first step.
In a new study in the journal Science, Eldred and colleagues described using this system to understand a fundamental question about vision that has remained surprisingly mysterious: How does color vision develop?
The researchers found that the blue cone cells, which detect blue light, develop first and that the red and green photoreceptor cells begin to develop later. They also found that thyroid hormone seems to be the critical signal in determining which light-detecting cells develop.
Ultimately, the researchers hope the insights could help develop treatments for diseases in which these light-detecting cells are depleted, such as macular degeneration. Better understanding of the process might lead to therapies for vision defects that develop in premature infants.
"The ideal goal would be to take a person's cells, convert them into stem cells, and then reprogram them and put them back in the person and treat whatever the disease is," said Robert Johnston Jr., a developmental neurobiologist at Johns Hopkins who leads the lab where Eldred works.
But organoids have limitations. Human retinas are about 10 times as big as the organoids, which top out at about the size of the inner ring of a piece of Cheerios cereal, Eldred said. Plenty of questions remain about how well they mirror eye development in the fetus, since they lack many other peripheral structures. But the ability to grow these organoids in a dish over a time scale that mirrors human development provides a rare window into questions that can't otherwise easily be probed.
"In the past, if we wanted to work out the developmental mechanisms underlying a particular process, we would turn to model organisms like mice or zebrafish," said Thomas Reh, a scientist at the University of Washington who studies the development of the eye. Reh, who was not involved in the study, called it "really excellent basic biology."