Dialing up an embryo: are olfactory receptors digits in a developmental code?

Are olfactory receptors digits in a developmental code?

William J. Dreyer wonders why cells don't get lost as an animal develops. He has long puzzled over how fingers and toes emerge from a growing limb, embryonic cells coalesce into a beating heart, and the billions of cells in a brain connect in just the right way.

He may have found the solution right under his nose.

In the Aug. 4 Proceedings of the National Academy of Sciences (PNAS), Dreyer lays out the provocative idea that the cell surface proteins in the nose that detect odors also help assemble embryos. He argues that these olfactory receptors and related proteins act as identifiers, much like the last few digits of a telephone number, that help cells to find their intended neighbors in a developing embryo.

"I've been searching for these last digits for 20 years," says Dreyer, a biologist at the California Institute of Technology in Pasadena. "No one can say for sure [the new theory] is true, but I'm up to 90 percent confident."

Dreyer's hypothesis rests on a precarious foundation: a small number of published experiments, a computer-aided analysis of genetic databases, and several unproven but plausible assumptions. Its bedrock is the 1992 discovery of the genes encoding the cell surface proteins believed to capture odorants in the nose. Mammals seem to have more than a thousand such genes. Even a simple worm has at least 550 olfactory receptor genes, comprising more than 5 percent of its genome, notes Dreyer.

Recently, investigators have made the surprising observation that these complex proteins--all of which crisscross a cell membrane seven times--play a crucial role in the development as well as the function of the olfactory system. For the nose to work, its sensory nerve cells must send out long extensions, or axons, to connect with the brain region called the olfactory bulb. This area begins the processing of odor information.

Each sensory cell appears to display copies of a single olfactory receptor. Although cells that are making a particular receptor are randomly distributed throughout areas of the nasal cavity, all their axons converge on one of two olfactory bulb regions specific to that receptor.

The olfactory receptor protein on an axon somehow determines where on the olfactory bulb it will hook up. Peter Mombaerts of Rockefeller University in New York and his colleagues have crippled genes for individual mouse olfactory receptors. The sensory cells employing those genes sent their axons toward the bulb, but the axons stopped well short of their targets. While other cues apparently guided an axon to the general vicinity of its target, the receptorless cell could not pick out its exact destination.

Dreyer now theorizes that the axon's olfactory receptor looks for copies of itself on cells in the olfactory bulb. "Maybe they're the same: the seeker and the target," he says.

From this speculation, Dreyer developed the idea that an axon migrates to its target along a gradient of different olfactory receptors to which it binds more and more tightly. Only when the axon meets a bulb cell bearing its own receptor, the one to which it binds most strongly, is its journey completed.

If the olfactory system resorts to such receptor gradients, perhaps so does the whole developing embryo. "If you have such an elegant system for building one part of the brain, are you really going to invent something totally different for the next piece of the brain, or for building a finger or heart?" says Dreyer.

Last year, Dreyer plunged into genomics, a fledgling field that employs computers to survey the flood of data on newly isolated genes. He began to examine databases of expressed sequence tags (ESTs), which represent fragments of genes that are active in cells. Searching through large EST databases, Dreyer found that ESTs from the liver, lung, prostate, eye, kidney, heart, testes, and other tissues match olfactory receptor genes. His survey, supported by several studies from other research groups, suggests that all tissues make at least a few olfactory receptors.

"What are they there for? They're not there to smell the roses," contends Dreyer. "They're there for the receptor gradients that pull all types of cells together."

The biologist emphasizes that other cell surface proteins, as well as proteins secreted by cells, would also help olfactory receptors assemble embryos. "These molecules fulfill many of the addressing functions... by providing the equivalent of country codes, area codes, and regional codes, etc.," he proposes in PNAS.

Dreyer acknowledges that his picture of olfactory receptors offering embryonic targets rests on the contention that, in addition to recognizing odors, these proteins can bind to copies of themselves and to similar receptors. "These things are built to recognize molecules," he says. "It's perfectly reasonable, [but] there's no evidence. It's pure hypothesis."

Dreyer has outlined several experiments to test his theory. For example, he encourages researchers to mark when and where in an embryo the individual olfactory receptor genes are active. Those experiments would have to be analyzed carefully, he notes, because an embryo may show only a speckled pattern of cells expressing a particular receptor, much as only a small number of people in a specific area code would have the same last four digits in their phone number.

Mombaerts suggests another test: Scientists could create mice in which a large number of olfactory receptor genes have been disabled.

While a few biologists have already dismissed Dreyer's hypothesis as farfetched, others are keeping an open mind. "The notion that olfactory receptor genes may be expressed in other cell types than olfactory sensory neurons deserves our full attention. I am afraid that this issue has never been seriously addressed," says Mombaerts.

John Ngai, who studies development of the vertebrate olfactory system at the University of California, Berkeley, admits he was ready to reject Dreyer's theory but found he couldn't.

"One experiment could tell us that the theory is totally impossible or implausible, but I haven't found that experiment yet. I don't know of any one thing that unequivocally says it's wrong," he says.

As for Dreyer, he has faced skepticism before and had the last laugh. In 1965, he and a colleague put forth the radical notion that immune cells shuffle DNA sequences to create the genes encoding the many antibodies and cell surface proteins that recognize infectious organisms. The idea was ridiculed initially but later proven correct. Dreyer is now hoping that history will repeat itself.