Evolution isn't simply about the genes you gain. It's also about the genes you lose.

The word loss has a painful, grieving sound to human ears, and so it can be hard to see how it can have anything to do with the rise of diversity and complexity in life. And until recently, evolutionary biologists didn't pay much attention to lost genes because they were preoccupied with the emergence of new ones. New genes, they found, can be produced in many ways. A gene can get accidentally duplicated, for example, and the copy can mutate, taking on a new function. Or pieces of two separate genes can get fused together, producing a new sort of protein. Or an old gene can get acquire a new switch that turns it on and off according to a different set of signals. As genomes of more and more species have been sequenced, scientists have combed them for new genes. They look for genes that are unique to a species, or some group of species, and are not found in distantly related organisms. They want to sort out the old genes common to much of life and the new ones that created a new body plan in one lineage. Consider the genome of one of the closest living relatives of vertebrates, a delicate sleeve of a creature called Ciona. Scientists found that over 2500 of its genes (a sixth of its entire genome) can also be found in the genomes of vertebrates such as fish--but not in the genomes of invertebrates such as fruit flies or vinegar worms. So here, scientists have argued, may be some of the genes that set us vertebrates apart.

But in just the last few years, evolutionary biologists have also been getting interested in the genes that have vanished. The mutations that erase genes are pretty well understood. A gene may initially get shut down by some disabling mutation. Later, through a copying accident, the gene may get snipped out of the genome altogether. These deletions can be devastating, causing swift death or long agonizing disease. But in some cases, the loss can be borne. Individuals manage to survive without the gene, and over time, more and more of them emerge, until the gene disappears from the species altogether.

Gene loss is particularly important in the evolution of the parasites and mutualists that live within our cells. We depend for our very survival, for example, on oxygen-consuming bacteria that invaded our cells some 2 billion years ago and became mitochondria. Comparisons to their free-living relatives have shown that mitochondria have lost the vast majority of their genes, holding only onto a few they still need to keep up their end of the symbiotic bargain they have with their hosts (us). Losing a gene can actually be an advantage to an organism that live in a host that has genes of its own that produce proteins that serve much the same function as its own. A relative of mitochondria has also stripped down, but for a different reason. Rickettsia, the cause of typhus, can only live inside cells, but it is a deadly pathogen rather than a helpful mutualist.

But free-living organisms have lost their own fair share of genes as well, and those who overlook it may misread the history of life. Case in point: Bacteria can acquire genes not just through heredity, as we do, but grabbing them from other bacteria. (Imagine acquiring someone's DNA through a handshake, your eyes turning from blue to brown. It's a bit like that.) Scientists have been debating how important these two routes of evolution have been for microbes. Do they trade just a handful of minor genes, or can they swap the very core of their genome?

Some new research suggests that much of the evidence for rampant gene trading may actually be an illusion created by lost genes. Think of a cookbook analogy. Imagine that some family in a remote village long ago developed a recipe for a blueberry soup. They keep the recipe a secret, handing down copies of the recipe only to their children. Over time, the children move to surrounding villages, taking the recipe with them and handing it down to their children. But gradually some branches of the family lose it, perhaps in kitchen fires or by accidentally tossing it in the trash. Many generations later, you take survey, recording who still has a copy of the recipe for blueberry soup. You find that most of the people who have it live near one another, close by the ancestral village. But there are also isolated families scattered here and there who also have copies of the same recipe. You might assume that in these cases, the rule of secrecy was broken, and members of the family handed out copies of the recipe to strangers. Only by understanding who had lost it, could you see that the rule had been upheld.

Bacteria have no monopoly on gene loss, though, as a new report in Current Biology makes clear. Australian biologists reported their study of the genome of a coral. Corals belong to one of the oldest lineages of the animal kingdom (a phylum known as Cnidaria, which also includes jellyfish and sea anemones). Cnidarians left fossils almost 600 million years ago, tens of millions before the first fossils of many other animal groups. They are also biologically simpler than most other animals. They lack brains or complex sensory organs, relying instead on nerves that form simple nets. They don't have a mouth and gut running from one end of their body to the other. Only after Cnidarians branched off on their own did new animals emerge with heads and tails, with different sorts of sense organs and neurons, with muscles for swimming and burrowing, and with many other tissues. The Australian biologists decided it would be interesting to compare the genome of a coral (Acropora millepora, the coral in the picture here) to the genomes of animals on younger branches of the animal tree. They compared its genome to ones from both invertebrates (fruit flies and vinegar worms) and vertebrates (humans).

Out of the 1376 genes that the Australian scientists looked at in coral, they found 492 matches in the other animals. But overall, these matches were far more like human genes than of the flies and worms. In fact, 58 of the coral genes (11%) could be found only in the human genome and were nowhere to be found in the other animals. In other words, a sizeable chunk of the genes that existed in the earliest animals have been lost in flies and vinegar worms, while they have survived in corals and humans. These lost genes may change the way scientists understand the evolution of animals. The researchers who used the Ciona genome to identify new vertebrate genes used only fruit flies and vinegar worms as points of comparison. In fact, a lot of these genes may not have all that much to do with the rise of vertebrates at all. Our search for what makes us special will have to turn elsewhere.

Why do species lose genes, and what effects do the loss have on their future evolution? It's puzzling, for example, that humans and corals still carry genes that must date back 600 million years or more--genes that are intimately involved in the development of embryos, for example--and yet fruit flies and vinegar worms (and presumably many other invertebrates) thrive without them. The geneticist Maynard Olson has proposed that losing genes isn't just a mutation organisms can learn to live with, as it were, but actually can offer a big improvement. According to his "less is more" hypothesis, losing a gene can open up a new ecological niche an animal's ancestors never could enjoy.

Olson points out that wild mice have a body clock that senses the changing length of day through the seasons, and uses that information to control when they can have babies. Lab mice have lost this clock, allowing them to breed year-round in their unchanging environment. "Less is more" could be the reason that many animals such as fruit flies and vinegar worms have lost so many genes. It may also one of the things that makes us humans unique. A comparison between humans and mice, for example, shows that 2% of the genes of our common ancestor that lived some 100 million years ago were lost. A closer look our immediate relatives--the apes--shows that we lack a particularly important gene, one that makes a molecule that studes the surface of their cells. It's particularly common on their neurons. And significantly, it appears that our ancestors lost the gene 1.5 mllion years ago, just around the time our brains began to expand dramatically. Scientists speculate that the presence of this surface molecule somehow held back the evolution of more complex brains. Only when it was gone could our ancestors explore their full evolutionary potential. Lose the gene, and you open up a new world.