Sex on Six Legs: Lessons on Life, Love, and Language from the Insect World - Marlene Zuk (2011)

Chapter 2. Six Legs and a Genome

SOME of the most cutting-edge discoveries about insect molecular genetics, and therefore about how genes do and don't dictate complex behavior, have been made because Gene Robinson was tired of harvesting fruit. As a student worker on a kibbutz in Israel, he was asked to "help out with the bees temporarily, and since I was bored to tears picking grapefruits, I volunteered. I remember I was smitten that very first day."

In his correspondence, he glosses over exactly why the bees were so appealing, but despite parental skepticism (he summarizes his mother's response as: "No doctor, no lawyer, where did we go wrong?"), Robinson went on to pursue a master's and later a Ph.D. in entomology. Now at the University of Illinois, he still professes an unabashed love of bees, which he has parlayed into one of the most compelling uses of genomics, the study of an organism's entire mass of DNA, anywhere in the world of biology. Robinson is interested in just how a complicated behavior such as the division of labor in a honeybee colony, where some bees go out and forage among the flowers while others stay home and nurture the young, is derived, first from the hormones coursing through the bee's body, then via the firing of nerve cells in the brain, and ultimately from the minuscule variations within a gene that directs the activity. He calls what he does sociogenomics, the molecular genetics of social behavior. It is where the genetic rubber meets the behavioral road, and it can best be understood using insects.

Before explaining sociogenomics, a bit of background about the new age of genomics, and about what we mean by sequencing a genome or having a "genome project," is in order. The genome is the total set of DNA in an organism, arranged into the chromosomes that are characteristic of each species; humans have twenty-three pairs of chromosomes, while cats have nineteen pairs, cows have thirty, silkworms have twenty-seven or twenty-eight, and a species of ant has just one. Sequencing a genome means determining the order of the four chemical bases that are the building blocks of the helix of DNA. The bases are called adenine, thymine, guanine, and cytosine, usually abbreviated with their initials A, T, G, and C. The genes themselves are particular sequences of the bases that contain instructions on the manufacture of proteins that make up the structure of the body or instructions on regulating when and how other genes become activated. Not all of the DNA consists of genes; scientists knew going into the Human Genome Project, the first of such efforts, that some amount of the material on the chromosomes would be noncoding, meaning it does not contain information about either gene regulation or the making of a protein. The genome sequence therefore consists of a long—a very, very long—string of four letters, grouped together in a particular arrangement unique to each species.

Once the Human Genome Project was completed in 2003, it was clear that more genomes needed sequencing. Many scientists wanted to put two animals next on the list. First would be the zebra fish, as a way to examine genes responsible for the development of a fertilized egg into an adult organism, and then the laboratory mouse, because as a mammal we could more easily compare its genes to those of people. Nobel laureate Sydney Brenner demurred, saying that "the mouse is too close. It hasn't had enough time to randomize, so you are confused by the commonness of origin."

What he means is that because we so recently shared a common ancestor with mice, our genetic material is already very similar to theirs. But which genes are the essential ones, the ones retained through hundreds of millions of years? How have genes changed to perform different functions? To answer that, we need insects. It's been 250 million years since the mosquito Anopheles gambiae and the fruit fly Drosophila melanogaster shared a common ancestor. That's roughly the same evolutionary distance that exists between humans and fishes, a third more than the distance between humans and chickens.

Of course, it's not an either-or situation. The zebra fish and mouse genomes have now been sequenced, along with those of the chicken, the African clawed frog, and a nematode called Caenorhabditis elegans. Genome projects are in progress for a whole host of others, including the European hedgehog, the green anole (a small lizard often sold in pet stores as a chameleon, although it is only distantly related to the true chameleons), and the gorilla, in addition to many invertebrates. Nevertheless, insects can reveal the process of evolution in ways that no other group of organisms can.

As I already pointed out, insects are the most diverse group of organisms on the planet—there are more kinds of insects than any other organism, they live almost anywhere except deep in the ocean, and they vary enormously in size, shape, food habits, and virtually every other aspect of life. A queen ant can live for decades in her nest, while tiny midges that circle over fast-running Appalachian streams can dispatch a whole adult lifetime, complete with finding a partner, mating, and laying eggs, in a prompt 45 minutes. That diversity makes it much easier to answer questions about the genes responsible for traits such as life span or body size, because we have so many different types of animals to compare. Even if we had genomes for all the primates, say, or even all the mammals, it wouldn't be as useful as having genomes for as many types of insects, because compared with insects, one monkey is pretty much the same as another when it comes to appearance and even behavior. A monkey is a lot more like a mouse than a grasshopper is like a flea. And of course insects are important vectors of diseases from malaria to typhus, as well as linchpins of our agriculture through pollination and pests because of their fondness for the same foods we eat. Without them, we cannot understand what makes life tick.

What's more, because we shared a common ancestor with insects so long ago, we can use them as a way to explore how we arrived at similar-seeming destinations with such radically different modes of transportation. For example, we are social and spend time and energy taking care of our young. Honeybees are social and spend time and energy taking care of their young, too. We share a fair proportion of genes with honeybees—but are the genes associated with social behavior the same in both of us? If they are different, how do they get similar results? If they are the same, why did the genes persist through evolutionary time in us and them, but not in thousands of other species?

Size, Junk, and Garbage

BEFORE exploring which insects have had their genomes sequenced and what those sequences tell us, it is necessary to look at a different kind of large-scale genetic information we can get for living things: genome size. Before we knew much about the chemical bases that comprise the DNA inside a cell, we could at least determine the amount of DNA itself. Indeed, calculating an object's size is one of the first things we do with something new, whether that something is a previously undiscovered mountain, a recently incorporated township, or a newborn baby (why the vital statistics of weight and length are so often included on birth announcements is a mystery, at least to me, but it testifies to our obsession with measurement).

Ever since the DNA molecule was discovered in the late 1800s, scientists were interested in the amount of it in different kinds of animals and plants. In the 1940s and early 1950s, the "DNA constancy" hypothesis, which stated that the nuclei of cells in various tissues contained about the same amount of DNA, and that this was roughly twice the amount contained in sperm cells, was used to test, and eventually support, the notion that DNA was indeed the source of genetic material.

Once this idea was accepted, it seemed plausible that the more DNA a species had, the more genes it possessed, and therefore the more complex it could be. Intuitively, people looked at genes like money in the bank; the more you have, the more you can buy. Scientists thus expected that smaller, simpler organisms such as amoebas or flatworms would have less DNA per cell than hamsters or birds of paradise. Much to their surprise, this turns out not to be the case. The amount of DNA—weighed in picograms, or trillionths of a gram—is not related to the apparent complexity of the animal or plant in which it resides. Knowing genome size is useful in deciding which organisms should have their genomes sequenced, for the purely practical reason that sequencing smaller genomes is cheaper.

Animals vary seven-thousand-fold in their genome size, and as you might expect, insects are champions of this variation. Among mammals, the smallest genome of 1.73 picograms resides in the Asian bent-winged bat, while the largest, in the red viscacha rat from South America, is really not all that much bigger, at 8.4 picograms. This size difference is dwarfed by insects, which vary 170-fold in genome size. Here the champion seems to be a mountain grasshopper, with the diminutive Hessian fly as its sparsely endowed counterpart. Humans, by the way, have genomes of a modest 3.5 picograms, which at least weighs in at more than the house fly, though less than that of the grasshopper.

Aside from the kind of Trivial Pursuit cum Guinness Book of World Records appeal of this kind of information (though, alas, clues about genome sizes are unlikely to come up in crossword puzzles), what does the variation in genome size—and its lack of relationship to the complexity of the organism in which it resides—mean? Obviously, more isn't better. Bluntly put by Ryan Gregory, a biologist at the University of Guelph and one of the world's leading genome size researchers, this decoupling of DNA content and complexity puts paid "the expectation that genomes consist of the genes, all the genes, and nothing but the genes."

So if the genome contains material other than genes, how did that happen? Furthermore, what exactly is that other material, and what is it doing in there? And why do some organisms seem to have so much more of it than others?

The answers to these questions are intertwined. Some of the "extra" material consists of free-floating bits of DNA, sometimes called transposable elements or, more colorfully, selfish DNA. These arise when a sequence of DNA copies itself several times and then just lingers as part of the genome. It is selfish because, a la Richard Dawkins's selfish gene, the elements benefit by making more copies of themselves, but they do not contribute to the functioning of the organism in which they reside. If there is no disadvantage to the organism of harboring them, or even if there is a cost but no means of getting rid of them exists, they will persist, cluttering up the genome and giving us those oddball genome sizes in some species.

Other noncoding DNA is often called junk DNA, which sometimes is used to mean all types of genetic material aside from the genes themselves, but more properly refers to copies of genes that used to be functional but are now obsolete. Like a manual lawnmower with a broken blade that you tuck away in the garage even after you've bought an electric model, the junk DNA clutters up the genome. In a distinction reminiscent of couples squabbling over organizing the closets, some scientists call DNA junk if it's not functional at the moment but could be useful at some hypothetical time, like that lawnmower, but garbage if it's not functional now and never will be, like—well, maybe it's best not to offer an example here. As with the transposable elements, junk DNA is thought to accumulate because DNA has an inherent tendency to copy itself unless otherwise halted.

Genome size is often, though not always, a reflection of body size, particularly among insects and other invertebrates. And insects that take longer to develop from eggs into adults have larger genomes as well. Another restriction on insect genome size seems to be the way that the species develops—does it go through a metamorphosis with egg, caterpillar, cocoon, and adult stages, like a butterfly, or does each successive stage look like a slightly pumped-up version of the one before, like a grasshopper? The butterfly types seem to have far smaller genomes than the grasshoppers, for reasons that are unknown. Also perplexing is a link between sperm length, which as I discuss further in a later chapter varies enormously among insects, and genome size. And intriguingly, all insects that exhibit social behavior, including not just bees and wasps but termites, as well as cockroaches that take care of their young after hatching, have reduced genomes, despite the vast evolutionary distance between these groups.

I look forward to the solutions to questions about genome variation, but what I like best about the measurements of genome size is the way they make our selves feel so literal, so concrete. Thinking about how many molecules can be crammed into a cell, imagining the adenines and thymines jostling for position, or the helices spooning like lovers in the nucleus, means that we can visualize who we are with startling clarity. Science writer Carl Zimmer titled his book tracing the history of our understanding of the brain and its relation to the mind Soul Made Flesh, in reference to the way that we can now see our essence in neurochemicals and gray matter. Thinking about the actual DNA, doled out in infinitesimal picograms in the genome, seems to make that translation even more tangible.

The Sequential Fruit Fly (and Mosquito and Beetle)

THE FIRST insect to have its genome sequenced was, as you might imagine, that sturdy workhorse Drosophila melanogaster. This was followed by the honeybee, a suite of other fruit fly species in the genus Drosophila, two mosquito species, the silkworm moth, and the tiny beetle that often inhabits the flour canister in your kitchen. More are on the way, and all are helping us understand the action of evolution on humans as well as our six-legged kin.

Let us begin with the fruit flies. D. melanogaster is the model species for genetic research, but other species of Drosophila lead lives that are both similar and different. Unlike the cosmopolitan D. melanogaster, D. sechellia, for example, lives only on the Seychelles Archipelago in the Indian Ocean, where it specializes on eating Morinda fruit, from a usually toxic plant. Drosophila grimshawi has elaborately patterned wings and is one of the extraordinarily diverse Hawaiian Drosophila, occurring only in a handful of remote locations. It is nearly a hundred times bigger than the puny D. melanogaster. A close relative of D. grimshawi, D. mojavensis is native to the Sonoran desert of the Southwestern United States and breeds on the spiky organ pipe cactus.

The flies were chosen deliberately to cover a broad range of evolutionary history; the different species shared a common ancestor anywhere from half a million to sixty million years ago. This is approximately the same distance between humans and lizards, all within a group of flies in the same genus. Many of the genes are similar in all of the species, but others are surprisingly different. Journalist Heidi Ledford referred to the "turmoil" of the genome that is visible only when genes are compared across species; genes appear and disappear, the time and place for them to be switched on and off is altered. Even those stalwarts the sex chromosomes had some surprises; some genes were thought to be expressed only in males because of their position on the X chromosome, but different species with the same gene did not always express it in the same way. The genes used to code for molecules that fight microbes—part of the fruit fly immune system—are much more variable than others, which makes sense given the rapid rate of change of the disease-causing organisms. Genes for detecting odors, crucial to animals that make their living and find their mates on fermenting vegetation, are also diverse. And in some cases, although the regulation of a pathway for making a protein clearly changed, the protein itself was still being produced, suggesting that so-called transcriptional rewiring might be commonplace.

In contrast to the desire to discover universal principles about the operation of genes from the workings of Drosophila, the scientists examining the two species of mosquito that have had their genomes sequenced had a much more practical motivation: they wanted to understand species that have such enormous effects on human health. The first species to be sequenced, Anopheles gambiae, is the principal vector of malaria in Africa. The second, Aedes aegypti, transmits yellow fever, dengue fever, and the less well known chikungunya virus; the latter was responsible for a recent outbreak in countries bordering the Indian Ocean that caused about 250,000 cases of illness and over two hundred deaths. Two U.S. Department of Agriculture entomologists, Jay Evans and Dawn Gundersen-Rindal, note that Anopheles was "the first animal to be sequenced, other than ourselves, whose actions have a strong direct impact on human lives." Although Aedes has a much larger genome than Anopheles, it doesn't encode many more genes, further supporting the idea that even closely related species can differ in the amount of noncoding DNA.

Once the sequence data can be used to identify functional genes, it should be possible to detect which genes are responsible for, say, successful transmission of the microorganisms that carry disease inside the mosquito's body, or for the mosquito's ability to use odor cues in sweat or exhaled breath to find a human to bite. The hope then is to tinker with these genes and breed a mosquito with a gut that is inhospitable to the malaria parasite, or one that cannot smell a delectably pungent victim nearby. Genes that are used to resist the effects of insecticides could similarly be altered to ensure that the mosquitoes remain vulnerable to certain chemicals.

If the fruit flies were sequenced to take advantage of the classic model system for genetics, and mosquitoes were sequenced in hopes of applying the knowledge to curing human disease, the flour beetle Tribolium castaneum could be said to have been sequenced because, well, no project in animal biology is complete without including a beetle. More kinds of beetles have been described than any other single group of animals—with over 350,000 species, one-quarter of all of the species of animals in the world is a type of beetle. The scientists who collaborated to sequence the flour beetle genome boast that beetles are "by far the most evolutionarily successful" multicelled organisms, and list, as if the insects were trying out for some kind of all-star reality television show, the many talents found in the group: "Beetles can luminesce, spit defensive liquids, visually and behaviorally mimic bees and wasps, or chemically mimic ants." I am not sure why these particular abilities are showcased, although there is a kind of "animals you might want with you on a desert island" kind of flavor to the selection. Interestingly, the beetles share with that other highly successful insect group, the ants, a lack of flight in day-to-day life; although most beetles can fly if necessary, their lives are mainly spent walking and tunneling on the ground. Whether this sacrifice of fragile wings is the key to their profligacy is not clear.

Tribolium itself is a good choice, among all those hard-shelled crawling candidates. Because it is easy to rear in large numbers in Petri dishes or other small containers, it has already been the subject of other types of genetic studies for many years. It is also an economically important pest in stored grains, which means that discoveries about its genome could reveal genetic Achilles heels to be exploited in its control, an urgent need since up to now it resists all kinds of insecticides that have been used against it.

Despite all the attention paid to the fruit fly Drosophila and its kin, it turns out that Tribolium is more of an "ur-insect," so to speak, than the fly—in other words, the flour beetle's genes seem to be less specialized and more like that of the ancestor of the entire class of insects than do the Drosophila genes. Over 125 groups of genes that the beetle has in common with humans, for example, don't occur in the other insects whose genomes had been sequenced as of 2009, suggesting that Tribolium has some pretty basic genetic material. In fact, nearly half of its genes are ancient, with counterpart genes occurring in vertebrates. This primordial nature means that it will be easier to determine how genes have changed through evolutionary time by comparing various groups to the Tribolium, and to determine which genes are responsible for general features of insect biology, such as metamorphosis or molting, and which are more idiosyncratic, say, those controlling the ability to make honey.

As is the case for genome size, and for that matter body shape and appearance, the genetic information from insect genome sequences is much more diverse than that obtained from vertebrates. A few constants appear, such as genes associated with detecting odors or those used to produce compounds that fight disease, but others are far more specialized. Silkworms possess about 1,800 genes that aren't seen in mosquitoes or fruit flies, including some used to make silk; although all insects and spiders use silk in some form or another, for spinning cocoons or dropping down from ceilings, the silkworms seem to have some additional genes exclusive to their lineage.

Of course, the first step after sequencing is to find genes with particular functions. Once that is accomplished, the opportunity arises for new, and sometimes diabolical, methods of pest control. Scientists are currently trying to use genetics to make insects pass on the instruments of their own destruction. A gene that is innocuous in the presence of, say, a particular antibiotic, but lethal otherwise, is inserted into an insect. The insect is then reared on a diet containing the antibiotic until it is an adult, when it no longer feeds, and is released into the wild. After the insects with the manipulated genes mate with normal members of the opposite sex, they produce offspring containing the gene—but those offspring are out in nature, where the lethal gene takes effect. Other even more clever methods are in the works.

As with genome size, studies of genome sequences confirm the presence of a hefty amount of noncoding DNA. One researcher refers to it as "dark matter," similar to the science fiction-like invisible stuff of outer space, which conveys both the mysterious nature of the substance and the almost peevish response that its discovery has elicited. We all seemed to have expected Mother Nature to be more thrifty in her allocation of genetic material, maybe saving that extra DNA, like leftovers at dinner. Shouldn't somebody have made another organism out of those bits and pieces of adenine and cytosine? Or maybe we just don't like the idea that it doesn't take many genes to make a whole complicated being; as Ryan Gregory says, "The strikingly low number of genes required to construct even the most complex organism represents one of the most surprising findings to emerge from the analysis of complete genome sequences." Somehow we seem to feel cheated by our own simplicity.

Of course, it's not that we are simple, per se. It's just that, once again, we are reminded that evolution is a tinkerer, using what's at hand to make its products. I like to think of the nuclei of our cells, not as perfectly tuned whirring machines, each gear essential, but as vast echoing warehouses of factories. Entire machines are outdated and useless, left rusted in a corner but never taken away and demolished. Others are jury-rigged out of pieces from older models and newer ones, rattling jerkily through their paces but ultimately manufacturing something useable.

The Social Genome

ALTHOUGH honeybees, like mosquitoes, are enormously important to human well-being, the sequencing of the honeybee genome was heralded not just because it might help us fight the mysterious decline of colonies throughout North America, but because bees are such extraordinarily social animals. Gene Robinson, who eschewed fruit-picking to devote himself to bees, thinks studying their genomes can show us how animals can become so integrated that they are often described as a single superorganism. According to the great biologist and ant lover E. O. Wilson, "If Earth's social organisms are scored by complexity of communication, division of labour, and intensity of group integration, three pinnacles of evolution stand out: humanity, the jellyfish-like siphonophores [creatures such as the Portuguese Man o' War], and a select assemblage of social insect species." Where does this high degree of interdependence come from?

One of the most surprising pieces of news from the honeybee genome project, published in 2006, was the relative paucity of genes associated with defense against diseases, compared with the other insects that have been examined. Given the crowded conditions of your average hive, one might imagine that pathogens would spread faster than colds at a preschool, which should select for highly vigilant immunity. One possible explanation is that the intense social behaviors of the bees, for example, the grooming and licking that individuals are always bestowing on each other, obviate the need for other defense mechanisms. It is also possible that honeybees, domesticated as they have been for thousands of years, will turn out to be an anomaly in this regard, a question that the sequencing of other social insect genomes should help settle. Two other startling results were the small total number of genes in the honeybee genome, and the apparent conservatism in the rate of the genome's evolution, compared with the mosquito Anopheles and Drosophila melanogaster, so that at least for some groups of genes, bees are more like vertebrates than those other insects. Contrary to what had been believed previously, in fact, bees seem to have arisen quite early in evolutionary history, branching off before the beetles.

Bees do have a lot of genes associated with producing and detecting pheromones, chemicals used to communicate with other individuals, which is not so surprising given their reliance on signaling within the colony, and they have some new genes that are associated with collecting nectar and pollen. But do they have special "sociality genes"? Several years before the honeybee genome sequence was completed, Gene Robinson noted that the difference between highly social insects such as the bees and solitary species such as Drosophila was likely to lie not in the creation of entirely new genes, but in changes in the way the same genes were turned on and off, or in the amount of product a given gene made. With some exceptions, this has turned out to be the case. Indeed, Robinson and his postdoc Amy Toth suggest that just as developmental biologists have discovered "modules" in body plans, with wings, legs, and arms produced from similar groups of genes in different animals, behavior can likewise be broken down into building blocks.

One of the most significant elements of insect sociality is the division of labor that Wilson cited above. Unlike other insects, or even virtually any other animal except for a few oddballs such as naked mole rats, in ants, bees, and termites queens do queenly things like produce eggs, males mate, and workers, well, work. Within the workers, different individuals often specialize on particular tasks, for example, going out and collecting food, or cleaning up the hive. This division, like the stratification of human industrialized societies, allows the colonies to be much more efficient. And the whole idea of sterile individuals that nonetheless labor for the group as a whole is a hallmark of sophisticated social organization. But what determines the destiny of any one individual?

It's arguable whether being a queen in a social insect colony is enviable or not, what with the continual egg laying and never getting outside, but the dogma used to be that queen honeybees were made, not born, via the feeding of royal jelly, a substance produced from glands in the heads of the workers that is given in lesser and greater amounts to different larval females. Adult bees, regardless of their social status, do not eat royal jelly. If you got a lot of royal jelly, the thinking went, you became a queen, while more modest amounts destined you for a short, chaste life among the colony proletariat. In the words of royalbeejelly.net, "Royal Jelly, the queen's food, makes the queen into a bigger animal with superhero powers," which I suppose is true if being capable of laying massive numbers of eggs is viewed as the insect equivalent of making yourself invisible. The association between upward mobility and royal jelly has given rise to a number of claims about the substance's ability to cure everything from asthma to wrinkles, though in a more sober moment surely someone has pointed out that bees suffer from neither.

But now it's turning out that at least for some social insects, you are not only what you eat, you are also the way you were born. In honeybees, different nutrients interact with the genome to switch some developmental pathways on and off, for a much more complex picture than had originally been supposed. In some ant species and at least one kind of termite, females bearing one version of a gene are more likely to be queens, while females with another version end up being workers. A particularly odd version of this genetic influence on caste occurs in harvester ants, in which two genetic lines coexist; queens belong to one line or the other, but workers are a cross between them. If a queen's eggs are fertilized by a male sharing her pedigree, the larvae become queens, but if the father is from the other line, the daughters become workers (recall that sons are produced only from unfertilized eggs, so they don't enter into the calculation). The difference between queens and workers can also be due not to a gene or genes being present or absent but to the regulation of those genes. A recent study of honeybees found at least two thousand genes that were present in both workers and queens were expressed differently in the brains of the two kinds of individuals, further supporting the idea that it's not just what you have but what you do with it—or what it does to you—that counts.

Queens may also specialize, with multiple reproductive females starting a nest together and then divvying up the duties like housemates, so that one goes out and collects food and the other stays home caring for the offspring. Alternatively, in the fire ant common to the southern United States and named for its painful sting, some colonies have one queen and others have two or more. The fatter queens go solo, whereas the burden is shared, literally and figuratively, in the nests ruled by multiple, lighter queens. Queen physiology, and the way the queens are treated by the workers, are both controlled by genes.

The role taken on by a worker had also been thought to be, if not diet related, at least environmentally determined, with all older worker bees, for example, doing more foraging and younger ones staying behind as "nurses." Now, however, the picture seems both more complex and more genetically determined. The age-related changes in tasks occur, to be sure, but altering the genes can change the workers' behavior, making them go out to forage at a younger age than they normally would. At the same time, foraging is influenced by social cues such as the age of other colony members and the type of pheromones given off in the hive, which in turn can feed back to the worker and change the hormones secreted inside the worker's body, further altering behavior. As in the queen-worker distinction, gene expression differs depending on the task the workers do. Hives with differences in genetic makeup also show different patterns of work. Most interesting, when a queen had mated with multiple males, the resulting blended family of workers was more efficient at making honeycomb, rearing the young, and flying off to collect pollen and nectar than were colonies started by a queen that had mated only once.

A group of scientists at the University of Sydney performed a clever experiment to examine the genetic regulation of reproduction in honeybees. Like humans, bees are affected by carbon dioxide gas, but unlike humans, queen honeybees respond to CO2 by increasing their ovary development, as if they had just mated and were getting ready to start a colony. In contrast, if a queen is removed from a hive, something the workers can detect immediately, the workers respond to the gas by suppressing their ovary development, just as if the queen were present and producing all the eggs (worker bees are able to lay eggs, although their sisters often prevent them from doing so).

The researchers, led by Graham Thompson, placed virgin queens and queenless workers in a chamber with CO2 for 10 minutes and then compared the gene expression in the bees' brains as well as their level of ovary development at intervals of several days. They examined twenty-five genes and found differences in expression in ten of them, suggesting that the bees are exquisitely sensitive to small changes in their environment and that the actions of their genes are altered accordingly.

Where did the extreme social behavior in these insects, with its self-sacrificial sterility, come from in the first place? The study of the honeybee genome, as well as detailed information on the genes of social and nonsocial species, supports an idea that had been around for a while among entomologists: start with mothers sticking around to feed their young, and go from there, progressing from maternal care to the more generalized care of siblings. Many insects show a more modest amount of social behavior than the ants or honeybees, as I describe in the chapter on parental care; they may guard their eggs, bring food to the developing young, or join forces with other females to rear offspring collectively, and they provide good test cases for this idea. Toth, Robinson, and a group of colleagues used the common paper wasp to see if care of sisters and care of young were governed by the same genes. Although the genome for the wasps has not yet been sequenced, the scientists used an innovative technique to characterize short segments of DNA that were already known to be associated with social behavior in honeybees. Although the bees and wasps last shared a common ancestor 100 to 150 million years ago, the genetic material that was examined turns out to be amazingly unchanged.

Paper wasps do not show the extreme differences among castes seen in ants or honeybees, but the scientists were able to examine DNA from four kinds of individuals. Foundresses are the females that start up a colony in the spring, usually by themselves, which means they forage as well as reproduce. They rear the first generation of daughters, who then become workers, allowing the foundress to become a queen and spend all of her time laying eggs. Finally, gynes are females that mate late in the year, spend the winter in a sheltered place, and then emerge in the spring to become foundresses.

Despite the identical outward appearance of the four types, and the fact that in some cases they are actually the same individuals performing different tasks, the researchers found that the wasp females differed markedly in the expression of genes in their brains. Workers had brains that were more like the foundresses that also cared for young than the queens and gynes that reproduce. Some of the genes that differed in expression were related to the production of insulin, an important component of nutrient regulation in insects, as in humans, which suggests that becoming social involved evolutionary changes in how food is perceived and processed. Toth and Robinson believe that the path from completely solitary to intensely social made use of a kind of molecular toolkit common to the ancestors of both kinds of behavior, modified in small ways as natural selection acted on the components. This differs from earlier ideas that new behaviors needed new genes.

The Collaborative Dictator

RESULTS such as these are leading us to a much better understanding of what it means to have genes control anything, whether that is social behavior or eye color. People often assume the existence of a gene "for" a trait, so that if you have the monogamy gene, for example, you won't cheat on your spouse, but if you lack it, your infidelity is inevitable. Studying genomes shows this is futile. First, genetic material is often redundant, nonfunctional, or just plain disassociated from any obviously useful protein. Second, genes are the great recyclers—all of our genes were modified from preexisting ones, with some new mutations that occurred at random thrown in. The genes associated with parental behavior are related to those that make a bee more likely to feed her sister, which are also associated with myriad other behaviors. This means that no gene can be associated de novo with a single trait and that trait only. Third, and maybe most important, genes are regulated with a complexity that is only just beginning to be understood. As in the paper wasps, it's not the genes themselves that change, it's the conditions under which they are expressed, and that regulation requires a host of other genes.

This is not to suggest that we shouldn't try to explore the genetic basis for behaviors such as courtship or maternal care. On the contrary, the new technologically sophisticated methods can reveal extraordinary detail about the mechanisms behind even complicated behaviors. But we should abandon, once and for all, the antiquated notion that we will ever have a catalogue of genes that can be neatly assigned to one and only one characteristic, that a gene associated with long eyelashes will have no truck with one making us more likely to prefer salty foods. Genes may dictate the production of proteins, but they do so in a maze of collaboration with other bits and particles of DNA.

What Next?

GENOME sequencing seems to induce a kind of greed in scientists, a hankering for more species with more variants of behavior and appearance. Many biologists have a favorite study organism and so often would love to have "their" animal or plant sequenced next. As the costs of processing samples decreases, the need to set priorities won't be quite so pressing, but right now several scientists have come up with justifications for "wish lists" to help guide future efforts.

Evans and Gundersen-Rindal used four criteria to evaluate groups of insects for their place on the list. First was genome size: smaller genomes are easier to sequence, and we already have an idea of genome size for many of the major categories. As mentioned earlier, flies, butterflies, and the bees and ants all have relatively small genomes, while grasshoppers and crickets, cockroaches, and silverfish, those odd little wingless pests in libraries, all have rather large ones. The central database called GenBank already has information on proteins in some of the groups, particularly flies, which also helps in starting a sequencing project. Evans and Gundersen-Rindal also ranked the insects for species diversity within each group, arguing that we would be better off working with diverse groups because they are likely to have more researchers working on them. Finally, they scored the insects for their effect on humans, where, as you might imagine, the elusive silverfish were pretty low on the scale. Overall, they plumped for more flies, more social insects such as bees and ants, and more beetles, with some moths and butterflies thrown in as well.

Beetles, particularly dung beetles, were also favorites of biologists Ronald Jenner and Matthew Wills, who suggested that the horned dung beetles in the genus Onthophagus would be particularly useful. As with antlers on deer and moose, the horns are more developed on males and are used in fights between rivals for females, allowing researchers to examine the genetic control of sexual differences. What's more, horn size is influenced by the environment in which a beetle matures, with better nourishment yielding more impressive weaponry; this could yield insights into the ways that genes are switched on and off by external factors.

Using criteria roughly similar to those of Evans and Gundersen-Rindal, myrmecologist and insect photographer Alex Wild mused about which ants would make the best candidates for genome projects. He settled on seven prospects, including the leafcutter ants of the New World tropics, which as their name implies slice off bits of vegetation that they bear off to the nest, where the material is chewed and used as a base for fungus gardens. The wood ants were another favorite, with many examples of social parasitism, potentially giving insight into the evolution of this unusual life history. One of the responses to Wild concurred with his proposal of another species, the bullet ant, which has an exceptionally painful sting, although the enthusiasm seemed to stem more from a desire for revenge by a victim than any biological justification.

The future clearly contains no shortage of animals to examine. I have a sneaking interest in those silverfish, though. Turns out they have some pretty bizarre mating tactics; the male spins a thread between a vertical object, such as a twig, and the ground and places a sperm packet beneath the thread. He then coaxes a female to walk under the silk, where she picks up the packet with her genital opening. After the sperm have drained into her body, she detaches the packet and eats it. The genetic story behind this kind of sex at a distance must be pretty amazing.