How to Build a Dinosaur (21 page)

Transplanting did not cause any continuation in the growth of the tail. In fact, the tails stopped growing, perhaps because removing the tip was enough of a wound to affect growth through physical damage to the tail, or perhaps because there was a group of cells at the tip that formed the organizing center for tail growth and once that was gone, transplanting of another tip was not effective.

The use of retinoic acid did seem to work. The tail did continue growing and adding vertebrae, but not in a way that was meaningful to the experiment. The retinoic acid, Hans said, “pushed tail growth to the upper range of normal development. So it had some effect, but it didn’t break it out of the cycle.”

Even if a full tail had grown, however, it would not have been proof that Hans had been able to reactivate an ancestral pattern of development. It might have been a freak of sorts, because, he was realizing, he did not have a good enough baseline model of how the tail grows and turns into a pygostyle in normal development. It became clear to him that the tail was a far more complex system than he had imagined, and that if he were to get more than a hint of growth, he would need to know more about the normal development of the tail and pygostyle in the chick embryo. The initial hint of growth was encouraging. He was obviously on the right track, but he needed to know much, much more in order to avoid creating something that looked like a longer tail but was simply an embryologist’s trick, finding a way to promote growth that was not an ancestral pathway.

Suppose, by throwing different growth factors at the tail at different times, a way could be found to grow a full-length tail. Unless he knew the normal pathway of development in detail, and could say exactly at which step he had intervened, and why this was likely to be the evolutionary change that resulted in no tail, all he would have achieved would be a circus attraction. As he put it, “One danger would be that we’re just sort of making a new anatomy. But I want to be sure that we’re playing with or manipulating the exact same system that is there normally.”

The pitfalls were clear. To take an extreme example, thalidomide given to pregnant women can result in babies being born with short flipper- or finlike limbs. But if these were to be claimed as atavisms—throwbacks to an earlier ancestor—the claim would be ludicrous. The same foolishness might misguide us into thinking we were looking at reverse evolution in the case of the tail. What we are hunting for is a change in the same system that changed during evolution.

Had the shotgun approach worked and a tail grown out to all eighteen vertebrae, Hans could have gone back to use the genetics and chemistry of that change as a starting point for digging down to the basic developmental genetics of tail growth. Or, if he were working on the limb, turning the wing into a forelimb, he could have gone to the scientific literature, which is wide and deep on limb formation. He would have found there enough work to create a baseline system of gene activation during limb growth and to know whether a change in limb growth was a modification of the baseline system that might be related to how it evolved, or whether it was simply a crude way of throwing off the course of development.

With the tail, he found, there was no such foundation of basic research describing how it grew. Research had been done on tail initiation but not on how its continued growth was mediated, and by what genes. He found he had to start from scratch and do basic developmental research. “We’re essentially having to try and map out these very basic genes and very basic research that has to be done just to make some sense of normal tail development.”

The tail is not a major identifiable part of some of the most studied evolutionary transitions—from fish to land, for example. Limbs have gotten quite a lot of attention. But, says Hans, tail growth is actually more complex than limb growth and potentially more profoundly important to development.

“The tail is, in essence, a little more complex than the wing or the leg because it has a lot more structures to it. The limb is just a bag of cells growing out distally and it brings with it the skeletal structure, the muscles, the nerves and blood. But that’s about it.

“The tail has all of that, plus it has two other things that are completely different from the limbs. It has the central nervous system in it, which is crucial. Secondly, it has the notochord. And the notochord is something that’s very, very basal within all chordates. All chordates have the notochord, and it’s another signaling center. The notochord, the central nervous system, and the surrounding mesoderm all set up a sort of cycling of events that helps to generate front-to-back segmentation and front-to-back polarity.”

This is about as fundamental a process of growth as there is, determining bow and stern. The whole tail system and its development may turn out to have major import in developmental biology. “Clinically, central nervous system, notochord, and axial patterning are really fundamental for developing animals. There are more defects known at that level of development than there are at pretty much any other sort of developmental hierarchy.”

What has happened is that the questions we asked have led Hans to take on a fundamental aspect of development that has not been studied in any detail. “We’re helping to reveal a whole new developmental system, which might be very fundamental because it involves the whole body axis, the front-to-back axis, and making it longer and shorter, and so the mechanism driving this or controlling that level of development could have very, very profound implications.”

Why, for instance, do we have so many vertebrae? As with so many other aspects of animal design, there is clearly no design going on. Instead we see changes built upon existing systems, and evolution can only be understood by seeing previous systems and finding out how they evolved.

If the basics of tail development can be understood, then it becomes possible to ask these questions, make hypotheses, and test them in the lab by manipulating the normal program of development. Hans is trying to get to the most basic level of gene regulation. Vertebrae are derived from body segments called somites. There has been a lot of study and a body of knowledge built up about how each somite is built and its boundaries defined. What Hans hopes to find out is what turns on the somite generation engine.

If that key can be found through mapping out this complete developmental system, then the work can turn back to evolutionary biology. If the key can be found that turns on and off the tail growth process down at the somite level, the genes can be sequenced, the process mapped out, and he will, he hopes, “find out exactly what’s changing on a molecular basis across very different-length animals—alligators and birds, for example.”

This would be a major breakthrough toward understanding the molecular basis of large-scale evolutionary change, “whole skeletal level changes across millions of years.” Nothing like this has been done yet.

Of course, this began with a paleontological question about birds and dinosaurs, based on fossils. And it is an interesting lesson from the point of view of interdisciplinary studies. A paleontologist says, why not mess with a chicken embryo so that it grows into a nonavian dinosaur? It can’t be that hard. After all, they are both dinosaurs. They share a very similar skeleton. In the grand scheme these are small adjustments to a basic body plan and such adjustments, we are assured by evo-devo, are the result of changes in gene regulation, not a complete new suite of genes.

So let’s see what regulates limb growth, the growth of teeth, and tail growth—that ought to be the simplest of all. Hans tackles the question, assuming at first that it should not be that difficult, and lo and behold, a hole in our knowledge of vertebrate development becomes glaringly obvious. We don’t really know, at a molecular level, what runs the growth of a tail, which, it would seem, is central enough that it could be said to wag vertebrate development.

I know Hans agrees, and I certainly think, along with other scientists who are pushing for interdisciplinary studies, that beyond the obvious value of molecular biology to paleontology, there is the value of having the fossil specialists work hand-in-hand with lab researchers. It is the fossil record that gives us the story of how life evolved and raises questions that can be pursued in the lab. The tradition of molecular biology is to look at the smallest changes in the greatest detail. The fossil hunters are the ones who understand the grand sweep of evolution. And it is good to think of us coming back into the fold of biology after having been given the cold shoulder for decades.

Still, it is a stumbling block and a delay in the plan to grow a dinosaur.

What Hans is doing now is developing a baseline model of how the tail grows. This will involve labeling cells in the growing embryo tail, using microinjections of dye to follow the pathways that cells take as the tail develops into the pygostyle. He needs to see where zones of growth and organization move as the embryo grows. And he needs to test for what is going on biochemically as growth occurs. So his students will test for concentrations of sonic hedgehog in the growing tail every four to six hours during normal tail growth.

They will also be testing the presence and activity of another family of proteins, the Wnt family, which includes Wingless (Wg) a famous protein and gene (the genes and the proteins they code for share the same name). The Wnt proteins are signaling factors that are involved in controlling patterns of growth. They were first identified in fruit flies but, like many of the most important factors, are found to be significant in all sorts of animals. In vertebrates, the Wnt pathway, as it is called, is active in limb growth and all sorts of other areas.

A great help to Hans in his attempt to identify which genes are active at different points in tail development is that, although the tail itself has perhaps not gotten the attention it deserves, the development of the chicken embryo has been more than thoroughly investigated. Researchers who study other aspects of chick embryology do so by staining the embryo at different points in its growth to show the activity of the genes they are looking for. They may be interested in the skull, ribs, forelimb, or hind leg, but chick embryos are small, and there is no point in trying to restrict their tests to one tiny area. So when they use a stain to show gene activity, the stain affects the whole embryo. The limb investigators do not look at the stain results in the tail, but they do preserve the images of the chick embryos. So Hans and his students combed through the literature to find indications of what genes were active in tail growth. Having identified the target genes and their protein products they will record in more detail their activity in sections of the growing tail.

Once a record is compiled for the degree of activity and the three-dimensional location of activity of sonic hedgehog and the Wnt family of genes at 104 to 156 stages in the growth of the tail (the work will be done by a postdoc and several independent-study undergraduate researchers), all the images will be fed into a computer and students will work to create a digital 3-D model of the growing tail.

“Then we can actually point to a few of those genes and say, that’s the one that we really want to hit aggressively.”

At that point, this first step in the attempt to grow a dinosaur will really begin, that is, the step of finding the gene or genes that stop regular tail development and create the pygostyle. It may seem ironic, given the complications, that Hans picked the tail as the easier system in which to create an atavism, easier than the wing. But the fact is that although the growth of the tail is very complicated, the action that turns that growth off may be quite simple, whereas there would be many detailed moments of turning on and off different genes to cause forelimb rather than wing growth.

Hans compares the situation to a mechanical one. “It’s kind of like the key to a car. You could turn the key on and the motor will run and produce all these patterns and rhythms coming out of it. Once you turn the key off, then it stops.

“And the key is relatively simple, compared to the rest of the car. I think that’s the kind of system we’re dealing with. Or I’m hoping.”

As for the dinosaur, even within the egg, Hans says, “The experiment I’m sort of envisioning at night is that you have a single embryo developing in the egg with multiple injection sites and multiple kinds of molecules, proteins, or morphelinos or things like this to be really fine-tuning the regulation of genes.” He continues, “So we’ll be able to inject different parts of the embryo at different times of development with different things. If we do that, if the timing and position are correct, we should be able to manipulate lots of different kinds of morphologies—feathers, wings, teeth, tails.

“It would just take a little bit of time to work out each one of those systems at very great detail, which we’re now doing for the tail. And other people are doing for the limbs for clinical work. And teeth are being worked out by other people for mammals and such, and then if we can just sit down and play with all these in concert, which has never been done before.”

The goal, in the end, would be to steer the embryo down the path it would have gone if it were something like a very early coelurosaur. If the classic genes in the chick embryo that have the codes for the proteins that are so essential for life and growth are very close to those of an ancestral, nonavian dinosaur, and if the changes, over more than 150 million years, have been almost all in regulation of the classic genes, then we could find the old pattern of regulation.

Metaphors can be dangerous, but we might imagine the development of an embryo as a story, a series of events in which every event determines what other events are possible. Just as in a novel, or the story of a life, every event has a consequence for the rest of the story. In life we don’t know what the consequences of choices will be. Will a date lead to marriage? Will a job work out? Will a chicken salad sandwich cause food poisoning? With the embryo we can do the story over and over with different choices in gene regulation, rewinding the course of development. We don’t have to give the embryo new genes, just adjust the growth factors and other chemicals that direct development. And by doing that we can see what must have changed during evolution, and what the old pattern of regulation was.