When Charles Darwin declared that variants within a species compete for survival, scientists did not immediately link this discovery to the development of the embryo.

Until now, that is. Harvard Medical School Assistant Professor of Pathology James S. Michaelson has written in a soon-to-be-published paper that Darwinian processes "provide the dominant force in development" and that "cell death is more a cause, than a result, of developmental organization."

Scientists have long wondered how a two-celled zygote develops into a complex organism such as a human. Michaelson's new findings seem to suggest that Darwinian processes of cell selection provide a basis for development. In other words, those cells best fit to survive embryonic development and life as an organism are chosen.

Most cell biology is currently dominated by Lamarckian evolutionary theory--which suggests that cells respond to changes in their environments--and the belief that cell death takes place according to a "blueprint."

Michaelson--who also leads a Dunster House seminar on the biology of the diseases of the developing world--began his research with an examination of the immune system and its ability to detect and fight foreign bodies, known as antigens. This highly specific system can destroy these antigens, which invade the body, while leaving a person's cells intact.

For a long time, researchers have accepted that natural selection takes place in the formation of the immune system. It was believed that huge numbers of white blood cells, which destroy antigens, were created, and that upon invasion, these cells compete to ward off the organism. Only a few are successful, and these multiply to create the millions of cells present in our immune systems.

Research Techniques

Extending this hypothesis, Michaelson began experiments on the liver cells of laboratory rats. Using flourescence microscopy, in which a chemical compound is "tagged" by a flourescent molecule, he showed that different liver cells, or hepatocytes, produced different proteins.

This flew in the face of popular scientific belief, which held that all hepatocytes were similar, and therefore proved that there is in fact genetic variation among liver cells.

Toxin Fighters

Upon further experimentation, Michaelson injected the rat livers with turpentine, a toxic substance. He found that he could thereby select for fibrinogen-producing cells, which fight the poison.

Based on these findings, Michaelson concluded that "changes in the overall synthesis of each plasma protein are mediated by selective growth and death among these specialized cells in the liver. In short, plasma protein synthesis fits our description of the Darwinian process, organized by cellular selection among a diverse collection of specialized cells."

Michaelson said that this idea of natural selection in the liver suggests the same process in the embryonic development of all organs.

"The growth of cells can be formative," said Michaelson. The mechanisms of development, he adds, produce a smaller number of cells than would be expected with regular growth rates. This process leads to the organization of cells within the organ, he says.

Michaelson said of his theory, "whether or not it will prove correct or useful remains to be seen."

"The similarities to humans are apparent," Michaelson said. "[The theory] could explain a lot of other things in development."

Scientists Skeptical

While reluctant to comment on a yet unpublished paper, other scientists interviewed said they are skeptical of the new theory.

"Cell death during development of the embryo is a well established phenomenon," said Bjorn R. Olsen, Hersey professor of anatomy. "It would be a radical idea--a catchy phrase, perhaps," he said of Michaelson's theory.

Olsen admitted that competition did exist locally between cells. Cells compete for growth factors, and those who have more receptors for the factors survive, he said.

But Olsen said, the term "battle for survival" is a bit drastic.

Similarity to Nervous System

Gregor Eichele, a researcher at Baylor Medical School in Houston, Tx., said that the theory reminds him of cell death in the central nervous system, which has been known of since the 1930s, he said.

But Eichele added, "I also feel that selection is not really random. Very specific regions undergo cell death, and they are always the same," he said.