How to Build a Dinosaur (20 page)

If we have proposed an evolutionary process like the one for feathers or for the change in digit identity, we could rerun that section of development, changing the developmental pathway as a means to testing the validity of our proposed evolutionary pathway. At last, we would have a truly experimental way of studying macroevolution.

So, why not grow a dinosaur? At least that’s the thought that came to my mind. Leaping over the many details, it seemed so obvious that if fairly small changes in development, which adjusted the timing and concentrations of growth and signaling, could have led to the evolution of birds from nonavian dinosaurs, we could readjust those changes in development and get a dinosaur. Thus my idea of growing a dinosaur from a chicken embryo.

I talked to a variety of scientists about the project, some in independent laboratories in Asia, who were ready to jump right in. One of the limiting factors was money. To take on the whole project at once I would have needed millions. The National Science Foundation does not provide grants of that size to paleontology. And, as is often the case, I had come up with an idea that didn’t quite fit into the standard categories of science. Millions to turn a chicken into a dinosaur might not be the most politically popular scientific grant. It was highly speculative, and likely to make people nervous about changing life in a fundamental way. Not to mention the issues of whether this was fair to the chicken or not.

I understand some of the concerns, although a number of them are, frankly more political than ethical. We do genetic engineering on mice all the time. By knocking out one gene or another we produce obese mice, diseased mice, crazy mice. None of these mice are going to survive outside of a laboratory environment, because most of the knockouts are disabling. So we don’t need to worry about invasions of fat, bald, schizophrenic mice. But the ethical principles are the same. Because these knockout mice are so important for studying basic genetics and for understanding human disease, the societal consensus is that the research is worth doing. Not everyone agrees, of course. There is a large animal rights movement, and some people within that movement are opposed to all experimentation on animals. So far they are a minority of the general population.

If someone were to achieve complete success in growing a dinosaur, so that I could present the adult animal along with the scientific paper at a meeting, one question would be whether there was a risk of such animals escaping. Could this be Jurassic Park? Well, one animal could conceivably escape, but it would at best have the chances of survival that a lone chicken would have. It certainly could not reproduce in kind, because we are only talking about causing changes in the growth of the embryo, not changing its genes. So the apparent dinosaur would still have a chicken genome. If by some miracle it did mate with a hen or rooster, depending on its sex, the result would be an old-fashion chicken. If it died, we could stuff it and roast it. It would taste, as the proverb says, like chicken.

Another question might be about cruelty to the animal. I won’t suggest that pretty much any life would be better than that of most of the billions of chickens that are eaten each year, at least not as a logical defense. But if we were able to do this correctly, we would have not a carnival freak, but a creature with a functioning tail and forelimbs, and useable teeth. If there were any indication that the chicken-dinosaur were in pain, we would not continue with the experiment. That would entail killing the chicken, but if we decide as a society that killing chickens is unethical, far more will have to change than us giving up an experiment. So, unless the chicken would suffer mental anguish at its dinosaur appearance that we could not detect, I would argue that the experiment would not be cruel to the chicken.

In any case, I did not try to convince Hans to try to grow a dinosaur because, although that is the goal I have in mind, and I hope to find a way to push, pull, or cajole researchers into that direction, there are many, many small steps to take first. Each of them is difficult enough.

Hans was already researching how the tail in birds first got shorter and then disappeared over the course of evolution. I thought, well, why not look at it from the other direction? Suppose we were going to go backward, to try to bring back the tail, what would we do then? I gave about forty thousand dollars from my own pocket to pay for a postdoctoral researcher to work on this problem. Hans is continuing to pursue the research on his own now.

What attracted him to the project was the chance to push paleobiology a step further into rigorous laboratory testing, and to push evo-devo a step further by creating experimental atavisms. In development an atavism is an ancestral characteristic that appears in otherwise normal embryonic growth. For instance, human infants are sometimes, although rarely, born with tails.

That seems to be an atavism, but there is always a question whether the characteristic is simply a defect or mutant form that, in its structure, looks something like we imagine a tail would look. Or is it a trait that we still have the genetic information for reappearing because of a change in gene regulation prompted by the environment? The medical literature includes references both to true tails, with muscles, nerves, and blood supply, and pseudotails, which are simply something that looks like a tail. All of these are small—a few inches long—and clearly an accident of development, even if they have muscles and nerves. They are not the long tails we might imagine on the primate ancestors of chimpanzees or humans.

An experimental atavism is something we would produce deliberately. If you can intervene in development of the embryo to produce an ancestral characteristic, that would be an experimental atavism. A very few experimental attempts to achieve this kind of thing have been done. Chicken embryos have been induced to grow teeth. In one case, mouse tissue was transplanted and teeth were produced, so this was not a true atavism, these were not the teeth that we would have seen in an ancestral bird or dinosaur. What that experiment showed was that the tissue of the developing mandible in the chicken was capable of responding to the signals for tooth growth.

In another experiment, however, changes in the presence of growth factors produced teeth in a chicken without any transplant of tissue. The teeth were consistent with those of archosaurs, the group that includes birds, nonavian dinosaurs, and crocodilians. If that report is correct, then the researchers achieved a true experimental atavism.

“The idea,” Hans said, “is that if there is an historical event, say an evolutionary transformation, that those events would not only follow some set of developmental rules, but some of those signatures of that particular developmental change or modification should be or may be present in the descendant forms.” The signature would be in the molecular biology of development, a moment, for instance, in the course of an embryo’s development when growth in one direction stopped, and restarted in a different way. The cause of this change could be found by testing for changes in the concentration of the different proteins (growth or signaling factors) that promote and direct development.

The disappearance of the tail seemed to be a good evolutionary moment that would have left a signature in embryonic development. Primitive birds, like
Archaeopteryx
, still had tails, but modern birds have lost them. It seems a good bet that this was a simple change that occurred, a turning off of the growth program that was keeping the tail going. Find the chemical switch, flip it the other way in embryonic development, and the result would be a bird with a tail.

Before he began an attempt to create an atavism, Hans had already done research on tail development and he had encountered several surprises. The fossil record showed that in birds long tails gave way to short tails, and then to no tails. It seemed that in the chick embryo one would find a short tail, with few vertebrae, beginning to grow, before it was stopped. But the tail was not a system that had been looked at closely in embryology. “There is no body of literature like there is on limb development for tail development,” Hans said, “because it’s a much more underappreciated structure and humans don’t have long tails.”

It’s not that we are, as a species, incurious about other species. Look at our fascination with dinosaurs. But money and attention for embryology tend to flow toward research that may have an application to medicine. We are likely to pursue with more energy and money an understanding of a devastating birth defect in spinal development, for instance, than what happens with tail growth, although as it turns out the two may not be so far apart.

At the very beginning of his research on the tail, Hans found a couple of curious things. First, and this was not the big surprise, the chick embryos were not starting out with a short tail. Although the chicken and other modern birds have only five vertebrae plus the pygostyle, the embryo started out with the beginnings or buds (anlagen) of eighteen vertebrae. This is in an embryo about an inch in diameter, the size of a quarter. “So there’s actually quite a long tail in these embryos,” Hans said. The growth of a tail begins in the embryo and develops in a clear way, “adding on more vertebral anlagen on the end.” But, he said, “Then I found that at a particular stage of development everything comes crashing to a halt.”

The growth of the notochord, which preceded the spinal column in evolution and helps organize the growth of the spinal column in the development of all vertebrate embryos, was disrupted. Instead of continuing to grow from front-to-back, and guide the growth of the vertebral column, at a certain point the cells at the growing tip of the notochord “look as if they’re disintegrating.”

“It stops, becomes disorganized, and then makes a ninety-degree turn.” This was the shocker. Tail growth is at least superficially similar to limb growth, and in limbs there is a group of cells at the growing tip that plays a major role in organizing and directing growth, and produces a lot of the proteins involved in promoting it as well.

“The tail seems to have something similar in birds,” Hans said. And, “It was known to be present in mice and zebrafish.” It was not present in another popular vertebrate,
Xenopus,
the African clawed frog, perhaps the ugliest of all lab animals, from a human point of view of course. “It seems to be just that
Xenopus,
being a frog, and from Mars, is doing something slightly different.” Frogs, apparently, grow by their own rules.

So what Hans found was going on as the chick embryo grew was that the group of cells (the ventral ectodermal ridge) that was conducting growth just disintegrated, and tail growth stopped. Immediately after the end of tail construction in one area, however, what seemed to be a second, similar area of growth, another ridge, started nearby, and the notochord, the scaffold on which the spinal chord and tail are built, took a ninety-degree turn.

“It stops, becomes disorganized, and then makes a ninety-degree turn toward this new structure, which is
totally unheard of.
” No one had found that before. “It’s at this point, when the notochord starts turning, that cells start condensing around the tip and start forming cartilage and then bone. This is the beginning of the pygostyle.

“This is something again unheard of. It has only been recorded in salmon, and salmon don’t have a pygostyle, and they’re not closely related to birds. So salmon and birds have done this very unusual trick of ossifying the end of the notochord. Nothing else does this. It might be a way to really stop tail development.”

These findings threw a monkey wrench into the evolutionary tree, the phylogeny of birds and dinosaurs, because Phil Currie, now at the University of Alberta, had described a dinosaur in 2000 that has what looks like a pygostyle. The dinosaur is an oviraptoroid, related to birds, but not directly ancestral. So either the pygostyle evolved twice, in different lineages, or the details of the dinosaur-bird lineage needed rethinking.

With this understanding of tail growth Hans and his postdoctoral assistant began in the winter of 2007 to try to make a chicken embryo’s tail grow. This was the beginning, in my mind, of growing a dinosaur, although that’s not what they were attempting. But it was the first attempt to modify the chick embryo in that direction.

They began snipping off the tip of a growing tail at one stage and stapling it, with fine tungsten wire, to a later stage, to see if the growth factors in action during early tail growth could override the stopping signals at stage 29. By convention scientists have broken down development into forty-six steps known as the Hamburger-Hamilton stages, after Drs. Hamburger and Hamilton, of course. The forty-six steps cover twenty-one days.

Presumably, a transplanted tip of a growing tail might direct an older tail to keep growing. If this work showed some effect, then the obvious candidates for signaling factors would be sonic hedgehog and fibroblast growth factors. The next step was then to add, before stage 29, retinoic acid, which is known to stimulate the release of sonic hedgehog, by injection or in microscopic beads. The hope was to keep growth going.

In either case, transplant or retinoic acid, if the tail were to keep growing, chemical probes could be injected to find out what genes were being expressed, what other growth factors were present. The probes are designed to find the messenger RNA that carries the instructions for the manufacture of growth and signaling factors. Messenger RNA is easier to find than the growth factors themselves, which are proteins. Because we know what the messenger RNA sequence is for sonic hedgehog, for instance, we can use a mirror-image stretch of RNA, called antisense RNA, that locks on to the right RNA stretches.

Hans and his postdoctoral assistant tried transplanting tips from stage 22-24 to stage 27-29, and then also from stage 17-19 to stage 22-24. They used up quite a few embryos, of course, but there is no other way to do these sorts of experiments. And in a world that eats eggs and chickens, the supply of fertilized chicken eggs seems almost infinite.