January 2003
In the 1830s, Charles Darwin first visited the remote Galapagos Islands, the now famous archipelago off the coast of Ecuador, where he noticed something striking. Of the dozen or so species of finch that inhabited the islands each seemed to occupy its own unique ecological niche. Hoping to explain his discovery, Darwin noted that the beaks of each species varied in size and shape. Darwin speculated that each finch must have evolved highly specialized beaks that gave them a survival advantage over other species within their habitats. This simple observation would fit well with Darwin's now famous concept of natural selection, the idea that each species must adapt to its environment to survive and prosper.
Today, "Darwin's finches" remain one of the most cited examples of natural selection, a cornerstone concept in modern biology. Yet, nearly 175 years since Darwin's seminal observation, researchers have managed to shake loose relatively few of the molecular details of how a beak develops. Nor do scientists have a good handle on how developmental adaptations could occur and produce beaks as structurally diverse as those of a toucan, an eagle, or a finch. The Inside Scoop spoke to NIDCR grantees Jill A. Helms, D.D.S., Ph.D., and her colleague Richard Schneider, Ph.D., both scientists at the University of California at San Francisco, who recently published on the subject in the journal Science. In the paper, Schneider and Helms succeeded in switching from duck to quail - and vice versa - a population of fetal cells that contribute to the formation of the beak. The result: The duck embryos developed short beaks reminiscent of a quail, and the quail embryos formed long, flat beaks similar to a duck. While this might sound on the surface like a case of "weird science," Helms said the experiments are anything but. She noted that the work sheds new light on how the face achieves its unique shape, or morphology. At the same time, because the human version of the transplanted cells orchestrates facial development, she said the knowledge gained from these experiments has indirect application to people. As Helms noted, by understanding the developmental programs that are hardwired into these cells, it may one day be possible to safely manipulate them to prevent craniofacial malformations in a developing child and/or repair facial wounds in adults.
Many people will be intrigued that you succeeded in switching the beaks of the two birds. But, upon a closer reading of your paper, wasn't your real focus the behavior of the so-called neural crest cells, the cells that you show are responsible for establishing the developmental pattern of the beak?
Schneider: That's correct.
So, let's start at the beginning. What are neural crest cells?
Schneider: They are a population of cells that are vital to the development of all vertebrates, which obviously includes birds and people. During the very early stages of development, neural crest cells emerge from the ectoderm, which is a primary embryonic layer that folds up and forms the brain and spinal cord. At this point, the cells start to look and behave like undifferentiated stem cells and soon migrate throughout the embryo. Depending upon its destination, a neural crest cell can differentiate into one of myriad cell types, including neurons, pigment-containing melanocytes of the skin, and, as mentioned in our paper, much of the skeletal and connective tissue of the head.
Why did you perform the experiments in birds? Why not mice?
Helms: Well, for a few reasons. One, it's possible to manipulate the embryo of a bird while it is still in the egg. Right now, it's not possible to perform these manipulations in mammals. Two - and most importantly - bird beaks vary in nature among species. Since we wanted to understand how shape and size emerge during embryonic development, the beak seemed like a good choice. We could detect differences in the form of the beak relatively easily. For example, we chose two avian species that have very different beaks because of their natural food sources. Quails eat small seeds, meaning they have small, short beaks that function like vices that crack open seeds. Ducks, which feed on water plants, have large, broad bills. If a quail was born with a duck's beak - and vice versa - there would be no mistaking it. And finally, there was the obvious parallel with Darwin's observations on finch beaks.
My understanding is neural crest cells appear when a chick or duckling is just a few millimeters in size. How were you even able to locate the cells in the embryo?
Schneider: We had a very good idea of where to look for them. Again, neural crest cells migrate extensively during fetal development, and scientists have successfully tracked where the various cells settle within chick embryos. We know the locations and developmental fates of most neural crest cells, including from my own series of transplantation experiments.
But how did you actually transplant them?
Schneider: It's one of those things that's easy to describe, but extremely difficult to do. You first create a small window in the eggshell that allows you to look down on the back of the embryo through a microscope. Using an extremely fine tungsten needle, you cut out the donor tissue that eventually will give rise to the beak. Then, very carefully, you suck this tiny piece of donor tissue into a pipette and lower it through a window in the recipient's eggshell. Having already removed the equivalent region in the recipient embryo, you replace the missing tissue with the donor tissue. I should add that we never let the embryos hatch. We let them grow in the egg until they reached the halfway point of their incubation period.
What was your hypothesis when you transplanted the cells?
Helms: We actually had two working hypotheses - and a number of people in the lab took sides on which hypothesis would prove to be correct. Half of the lab thought that the neural crest cells from the duck were preprogramed to make a duck beak, and the same would hold true for the corresponding cells from the quail. The other half believed that the influence of any quail-specific signal would be overridden by the duck host.
Scientifically, though, the first hypothesis - the idea that neural crest cells are pre-programmed - isn't necessarily well established. Is that correct?
Schneider: There are two schools of thought on cranial neural crest cells. One holds that these cells behave like stem cells. That is, transplant them into a foreign environment, and they will make whatever that new environment prompts them to make. A second school of thought holds that neural crest cells already have pre-programmed in them information about how to form the three-dimensional pattern of a tissue. Transplant neural crest cells into a foreign environment, and they'll just go ahead and make whatever they've always made.
Does your paper settle the debate?
Helms: No, but I think our work sheds new light on the debate. We found that the transplanted cells differentiated according to their original species, which is certainly significant. But we also found that differentiation involves a community effect, meaning it requires a number of neural crest cells from the same species to be in one place together.
A community of cells then is greater than the individual cell?
Schneider: Right. Cells operate mostly by responding to signals from other cells. In order for the neural crest cells to elaborate their inherent patterning program and therefore create the three-dimensional structure of the beak, they had to engage in crosstalk with neural crest cells of like origin. In those cases in which we transplanted fewer and fewer neural crest cells, the beaks became more and more like those of their hosts.
You mentioned the two schools of thought. It sounds as though the answer to the debate might possibly lie in the middle?
Helms: I think that this is almost always the case. One of the fundamental issues that this work brings to light is developmental flexibility of a neural crest cell seems to depend upon the part of the face in which it resides. For example, neural crest cells that colonize the throat region, known as the hyoid arch, seem to follow some sort of pre-programmed pattern. Other work shows that neural crest cells located in the lower mid and upper face are more plastic, meaning they are more responsive to signals within their environment.
Our work suggests that the three-dimensional architecture of the face involves an interplay between the neural crest cells and their environment, which is usually described as the epithelium that surrounds the neural crest cells. There are three types in the craniofacial region: the neural, facial, and pharyngeal epithelium. So, there are a lot of sources of signals that might influence the fate of the neural crest cell.
Your study demonstrates the type of cell that controls the size and shape of the developing beak is the neural crest cell. One possibility is neural crest cells are particularly prone to mutations, which would explain the diversity of beaks and faces in nature.
Helms: In theory. But I think this is probably one of the most interesting aspects of the work. That neural crest cell could actually be thought of as a conduit through which change occurs, and that has both positive and negative implications. Obviously, changes that allow animals to adapt to environments are good for the species. But change can be bad in the sense that it leaves these birds quite vulnerable to environmental onslaughts, too. So, I think that is a very important part of understanding neural crest cells.
Where does your laboratory go next?
Helms: Well, I think it is nice to be able to say that a certain population of cells influences the ultimate morphology of the structure. But it begs the question: How do these cells do it? We now have only a glimpse of how we think it is happening.
So, in essence, you have a target in sight now?
Schneider: Correct. We showed in the paper that one of the ways the neural crest cells change facial pattern is by changing the time course of when genes are turned on and turned off. So, in effect, neural crest cells caused the earlier induction of genes in the duck embryo, and, by so doing, formed a quail-shaped beak. But that still leaves you sort of hand waving when it comes to understanding how these altered patterns of gene expression ultimately affect morphology. One has to start thinking about linking up gene function with the actual shape of a skeletal element.
What might this work tell us about craniofacial development in people?
Helms: That to me is one of the most important issues. In addition to learning how these cells behave and might influence the incredible variations in the human face, there's also the issue of facial patterning that goes awry. Is there anything that we can do to treat these malformations? Today, conventional surgery is the only option. But, nowadays, the idea of fetal - or in utero - surgeries is not so far fetched. As the technology and the skill of the fetal surgeons improve, they will start to look to scientists who study these questions and ask if there is some gene product that we can replace to mitigate these conditions.
Another possible application is to exploit the stem cell-like qualities of these cells. It's possible that a neural crest cell, when given the proper environmental cues, will differentiate according to its new environment. This raises the question of whether these cells are found to some degree in adults? Or, whether they are present only during fetal and early post-natal development? At this point, we don't know the answer. Our hypothesis - and we have some data to support it now - is that these cells do exist in adults. What we are now studying is whether they have the potential to contribute to reparative processes in the head. The idea is, if adults possess these cells and they retain their stem cell-like qualities, they can be used to repair skeletal injuries.
But this is just the start, and there's a lot more to learn.
Schneider: Absolutely.