The Whale and the Antibody

Evolutionary biologists face a challenge that's a lot like a challenge of studying ancient human history: to retrieve vanished connections. The people who live in remote Polynesia presumably didn't sprout from the island soil like trees--they must have come from somewhere. Tracing their connection to ancestors elsewhere hasn't been easy, in part because the islands are surrounded by hundreds of miles of open ocean. It hasn't been impossible though: studies on their culture, language, and DNA all suggest that the Polynesians originally embarked from southeast Asia. We may never be able to retrieve the full flow of history that carried people thousands of miles to the middle of the Pacific, but we can know some things about it, and we can rest assured that some things are definitely not true (such as the sprouting-from-the-ground theory).

Whales are a lot like Polynesians. All living species of whales look a lot like each other, and not very much like any other animals. They all have horizontal tail flukes, blowholes, and smooth skin free of scales or fur. Darwin argued that whales were not simply created in the oceans in their current form, but instead descended from land mammals which had adapted to life in the ocean. He pointed out that whales share a number of traits with land mammals, such as milk and a placenta. Their blowhole connects to a set of lungs very much like those of land mammals and nothing like the gills of fish.

Darwin wigged out more than a few people with this argument. Whales just seemed too different, too distinct to have evolved by small steps from a four-legged ancestor. And creationists loved to point out how unlikely this transition seemed--on par with turning a cow into a shark. They also liked to point out that no intermediate fossils had ever been found. But as I wrote in my book At the Water's Edge, paleontologists began to find those fossils in the 1980s. Today, the transition whales made from land to sea is wonderfully well documented. Paleontologists have found complete skeletons of creatures such as the 45 million year old Ambulocetus (reconstructed here by the gifted artist Carl Buell). The transformation was not some sudden macromutation, but a gradual series of changes over millions of years, featuring shrinking legs, lengthening tails, loosening hips, and migrating noses.

In the coming century, I suspect fossils will help scientists reconstruct other major transitions. But they'll also start reconstructing others that have left no record in rocks A fascinating case in point has been published on line at the Proceedings of the National Academy of Sciences. Jan Klein and Nikolas Nikolaidis of Penn State have drawn a rough map that charts the evolution of the immune system.

Our immune system is as awesome as a whale's body--in terms of the complexity of its parts and the way those parts work together so well. It keeps viruses, bacteria, tapeworms, and even cancer cells at bay, while generally sparing our own tissues from its withering attack. All animals share a rudimentary immune system, but Klein and Nikolaidis focused on a second system that is found only in vertebrates. Only we vertebrates have immune systems that can learn.

This learning system is a network of cells, signals, and poisons. Among its most important cells are T cells and B cells. They originate in the bone marrow, although the T cells have to finish their development in the thymus, an organ near the heart. These cells are unusual in many ways, most important of which are some of the receptors they make on their surface. The cells have a special set of tools that cut up the receptor genes and paste them into new arrangements, so that the genes produce receptors with new shapes. Depending on its shape, a receptor can grab onto certain molecules. Those molecules may come from a bacteria toxin, or they coat nerves or muscle cells. Our bodies can usually eliminate the immune cells that have an affinity for our own tissue. If they don't, we end up with autoimmune diseases such as muscular dystrophy.

The surviving B cells and T cells are introduced to molecules from invading pathogens (antigens) by other immune cells called macrophages. The macrophages devour bacteria or virus-infected cells and then put some of the molecules of their victims on their surface. They travel to the lymph nodes to show off their conquests. If T cells or B cells bump into one of these macrophages, their receptor may fit reasonably well onto an antigen. That fit sends a signal to their DNA, triggering them to multiply. Some of the cells they produce receptors cut and pasted into newer shapes, some of which do an even better job of fitting on the antigen. These winners get to reproduce more. In other words, our immune systems use a version of natural selection to fine tune their recognition of pathogens.

These B cells and T cells can then fight off a disease. The T cells may destroy cells infected with the pathogen, because most cells in the human body have receptors they can use to display antigens. In other cases, they can whip up macrophages into a furious frenzy of killing. Or they may spur B cells to produce antibodies. The B cells spray out the antibodies into our bodies, and when they come into contact with their particular pathogen, they may drill into it, stop it from invading cells, or tag the pathogen to make it an easier target for macrophage attack. Some B cells and T cells that can recognize a pathogen sit out the battle. If we should be exposed to the same disease years later, these memory cells can leap into action so quickly that the new infection may not even make us sick.

You can find this same remarkable system in humans, albatrosses, rattlesnakes, bullfrogs, and all other land vertebrates. You can also find it in most fish, from salmon to hammerhead sharks to sea horses. There are some variations from species to species, but they've all got B cells, T cells, antibodies, thymuses, and the other essential components. But you won't find it in beetles, earthworms, dragonflies, or any other invertebrate on land. Nor will you find it in starfish, squid, lobsters, or lampreys in the water. All these other animals rely instead on rudimentary immune systems that cannot learn.

For those who reject evolution, this sort of pattern tells them nothing. Like everything else in nature, they can only wave their hands and declare it the inscrutable work of a designer (lower case d or upper case D as they are so inclined on a given day). But immunologists and other scientists who actually want to learn something about the immune system find this view useless. Instead, they look at how animals with an antibody-based immune system are related to one another. And what they find is both straightforward and astonishing. All of the living animals with an antibody-based immune system descend from a common ancestor, and none of the descendants of that common ancestor lack it. That means that the antibody-based immune system evolved once, about 470 million years ago.

I need to back up in the history of life a few hundred million years to explain how scientists know this. Studies on fossils and genes agree that everything we call an animal (including sponges and jellyfish) shares a common ancestor not shared by plants, fungi, or other major groups of organisms. Exactly when that ancestor lived is a subject of fierce debate, but one of the latest estimates puts the date at about 650 million years ago. This ancestor probably had a simple immune system, because all animals, from sponges on, have at least some sort of defense against pathogens. Over the next 100 million years or so, the major groups of animals branched off from one another, and while some branches evolved some new defenses of their own, the antibody-based immune system only appears in our own branch, the vertebrates.

Animals with some--but not all--of the key traits of vertebrates, such as heads and brains, lived at least 530 million years ago. The only living relics of these early branches are hagfish. Later, our ancestors also evolved a vertebral column, becoming true vertebrates. Lampreys represent the deepest branch of vertebrate evolution, splitting off perhaps 500 million years ago from their common ancestor with us. They lack many traits that other vertebrates have--most obviously a jaw. A number of other weird jawless vertebrates filled the oceans between about 500 and 360 million years ago, but except for lampreys, they're all long gone. One of these branches gave rise over 470 million years ago to fish with jaws--known as gnathostomes. Gnathostomes later gave rise to sharks and other "cartilaginous" fishes, as well as ray-finned fishes, and land vertebrates.

You may have already guessed the kicker of all this history. Lampreys and invertebrates don't have an antibody-based immune system. Sharks, ray-finned fish, and land vertebrates do. Sharks, ray-finned fishes, and land vertebrates all share a common ancestor that is not shared by lampreys or other invertebrates. The simplest way to explain this coincidence is to conclude that the antibody-based immune system evolved after lampreys branched off from our own lineage, but before sharks and other living gnathostomes began to branch apart. We can't dig up fossil antibodies, but we can know when they evolved.

Scientists have sometimes treated the transition from rudimentary immune system to antibody-based immune system as a great leap. Lampreys don't have antibodies, B cells, T cells, thymuses, or the rest, and all gnathostomes do. Some creationists have even tried to turn this into an argument against evolution, claiming that something as complex as the adaptive immune system could not have emerged gradually. But it's important to bear in mind that tens of millions of years of evolution separate our common ancestor with lampreys and the earliest gnathostomes. And in their new paper, Klein and Nikolaidis argue that the evolution of the antibody-based immune system was a lot like the evolution of whales: a gradual, step-wise process.

Most of the components of the antibody-based immune system were actually already in place long before gnathostomes evolved. Lampreys, for example, don't have a thymus, but they do have the structures and cell types that form the thymus. In gnathostomes, the thymus develops as cells switch on special genes in a particular order. Lampreys have these genes, as so many other animals. Instead of building thymuses, they build other structures, such as eyes and gill arches. It would have only required altering the switches that determine when and where these genes become active to produce a new organ.

B cells and T cells are known as lymphocytes. Lampreys don't have lymphocytes, but Klein and Nikolaidis point out that they do have "lymphocyte-like cells." (The picture above shows what these cells look like.) Lymphocyte-like cells develop like lymphocytes, under the control of many of the same genes that control the development of lymphocytes. Once they are mature, these cells have almost the same structure and chemistry as lymphocytes--but they don't produce the antibodies and receptors of B cells and T cells. Exactly what they do in lampreys isn't clear.

What about those receptors and antibodies? Klein and Nikolaidis point out that they aren't quite as novel as they may look at first. They are made up of building-blocks of simple proteins arranged in different ways. And guess what--many of these simpler proteins are found in lampreys and invertebrates, where they serve other functions. The same goes for many of the proteins that B cells and T cells use to communicate with one another. Other proteins are made by genes that are unique to gnathostomes, but show a kinship to entire families of genes found in other animals. The most likely explanation is that an ancestral gene duplicated by accident, and later one of the copies was recruited to the evolving immune system.

Klein and Nikolaidis point out that some truly new things appeared as the antibody-based immune system emerged. But just because something is new doesn't mean that it couldn't have evolved. The best-understood example of a new feature is the cut-and-paste machinery that allows B cells and T cells to mix up their receptors into new shapes. Scientists have been working out its evolution for years now, but just last week some scientists from Johns Hopkins published a paper in Naturethat brought the picture into remarkable focus. Our genomes are rife with virus-like sequences known as transposable elements that produce enzymes whose sole function is to make copies of the transposable DNA and insert those copies somewhere else in our genomes. In a few cases, these transposable elements have evolved from pests to helpers, carrying out important functions in our cells. The genes that are responsible for cutting and pasting immune cell receptors bear a clear resemblance to transposable elements in other animals. So the evolution of a new cut-and-paste mechanism was actually just the domestication of an in-house virus.

I suppose that creationists might claim that these components could not possibly have come together into an antibody-based immune system. But there's no proof behind this sort of categorical dismissal, just a personal feeling of disbelief. These folks would still be left with the fact that the evolutionary tree of life and the biochemistry of vertebrates and other animals are all consistent with a gradual evolution of this system. It would all have to be a spectacular coincidence, or perhaps an intentional deception on the part of the designer. Who knows. Who cares, really? (Aside from certain Pennsylvania senators.) What's exciting here is the future research that could shed more light on this transition. Klein and Nikolaidis propose introducing lamprey genes into vertebrates and vice versa to see just how close the ancestors of lampreys had gotten to an antibody-based immune system before they branched off on their own. Obviously, some half a billion years of independent evolution will muddy up the results, but it should be possible to see whether gnathostome immune genes can organize the lamprey immune system to act more like our own. What I'd be even more excited by would be a deep-sea discovery of a living fossil--a jawless fish that is more closely related to us than lampreys. They filled the seas 400 million years ago, and perhaps a few are lurking in some deep sea trench. Such a fish might have a crude antibody-based immune system, with only a few genes recruited and others yet to be pulled in. Perhaps it could do a mediocre job of learning to recognize diseases--but a mediocre job is better than no job at all.

It may sound like a crazy dream, but then again, so did walking whales.

Update 1/2/05: Panda's Thumb has more on the evolution of the immune system.