Reef relations: DNA shared by people and coral sheds light on animal evolution.

The reef-building coral Acropora millepora does not have a lot on its mind. In fact, it doesn't have a mind at all. The invertebrate has only a diffuse net of nerve cells, one of the simplest nervous systems of any animal. Thus, it shocked Australian geneticist David Miller to find that the coral's DNA contains genetic sequences corresponding to genes that guide the patterning of the incredibly complex human nervous system. Worms and flies don't have these genes, so he and other researchers had taken it for granted that the genes were relatively recent innovations that had evolved in vertebrates.

The nervous system genes are among a surprisingly large number of genes shared by vertebrates and A. millepora, but not by the worm Caenorhabditis elegans or the fruit fly Drosophila melanogaster, Miller and his colleagues have found. This discovery demands rethinking of the common ancestor of corals and all other animals, the researchers say in the Dec. 16, 2003 Current Biology.

Despite the presumed physical simplicity of this ancestor, "it must have contained many more genes than we had previously assumed," says Miller, who works at James Cook University in Townsville, Australia.

Evolutionary biologist John Finnerty of Boston University agrees. The ancestor must have exhibited "a stunning degree of genetic complexity.... It is extremely important to reconstruct the genome of this ancestor, since it gave rise to almost all of modern-day animals," he says.

More than 500 million years ago, cnidarians, which include corals, jellyfish, and sea anemones, began to flourish in the oceans. Only sponges are thought to predate them among the true animals.

Seeking to establish A. millepora's place in this evolutionary history, Miller and his colleagues performed a genetic analysis on the coral's larval form. They identified DNA sequences known as expressed sequence tags (ESTs), which derive from active genes in tissue. All told, the scientists came up with almost 1,400 distinct coral ESTs.

Next, they scanned databases of other creatures' genes for DNA sequences that match the coral ESTs. The vast majority of the coral ESTs correspond to DNA sequences shared by all multicellular animals. However, about 12 percent of the coral ESTs had a corresponding human gene but no match in the worm and fly DNA. Until this finding, those human genes were presumed to be specific to vertebrates.

"The study makes clear that many genes previously thought to be vertebrate innovations were in fact invented long before the origin of the vertebrates," says Finnerty.

Just 1 percent of the coral ESTs matched worm and fly DNA without also corresponding to a human DNA sequence, Miller's team notes. And when a coral EST was close to matching DNA sequences in all three animals, the investigators found that the coral's DNA tended to be more similar to the human DNA than to that of the worm or fruit fly.

The finding doesn't suggest that corals are more closely related to people than worms or flies are, cautions Miller. Instead, he explains, the same short reproductive cycles that have made C. elegans and D. melanogaster popular for laboratory research have enabled the two animals to diverge much more than coral and people do from the ancestor they all share.

GETTING LOST Because many coral DNA sequences don't have worm or fly matches, Miller and his colleagues conclude that C. elegans and D. melanogaster have shed many of their original genes during their rapid evolution.

The new work "is important in showing massive loss of genes in some animal lineages," agrees Eugene Koonin of the National Center for Biotechnology Information in Bethesda, Md. Similar evidence of substantial gene loss in worms and flies had already emerged from comparisons of their genomes with those of several yeast species, he adds.

Miller advocates completely sequencing the DNA of A. millepora, especially now that other researchers are deciphering all the DNA of a sea anemone.

Overall, Miller argues, evolutionary biologists must study the DNA of a great variety of creatures before they can truly get a handle on the evolution of animals. "We need a more representative range of genomes" than have been sequenced so far, he concludes.

Miller and his colleagues are now endeavoring to identify the functions of the genes that the coral share with people but not worms or flies. Miller hypothesizes that in coral these genes are limited to a single role. In contrast, many human genes and their proteins may have evolved to fulfill multiple duties, thus allowing greater complexity in the organism.

"We can learn a lot about the ancestral roles of these multifunctional genes from studying animals like coral," predicts Miller.