Gene Linked to Beak Length in Darwin Finch

Simple Variation Yields Coordinated Change in Complex System

The beaks of Darwin’s finches occupy iconic status in biology. A diagram of the exquisitely adapted variations in beak length, height, and width now seems like a glaringly obvious illustration of evolution in action.

But for Darwin, who devoted much more time and ink to his pigeons, the finch beaks were a retrospective revelation, realized after his return from the Galapagos Islands, and only when an ornithologically informed friend told him that his bags of different bird specimens were mostly finches.

Photo courtesy of Arkhat Abzhanov

On the Galapagos Islands, Arkhat Abzhanov visited the nests of up to seven species of finch daily. With no natural mammal predators, the fearless birds would sometimes perch atop Abzhanov’s head.

After Darwin pointed out the “diversity of structure in one small, intimately related group of birds,” in his own words, it took several subsequent generations of scientists to document that the beak sizes and shapes provide a rich case study of how natural selection can drive extremely rapid evolution of life in response to changes in the environment. Now, researchers in the lab of Clifford Tabin, HMS professor of genetics, have added a new chapter to the classic story.

The team has discovered the first genetic and molecular underpinnings of the different beak morphologies. Two years ago, they reported that a gene known to shape the faces of laboratory animals also sculpts the height and width of the finches’ upper beaks.

Now, the researchers have found another gene whose activity independently contours the third dimension, length. In addition to identifying a new player in craniofacial development, the results provide insight into the nature of variation, the indispensable raw material for natural selection to act on. The findings are reported in the Aug. 3 Nature.

Earlier and greater expression of the gene for bone morphogenetic protein 4 (BMP4) forms the deeper, broader beak, ideal for cracking seeds. The researchers first observed this in finch embryo samples and then demonstrated its direct effect in the chicken embryo model system. An earlier and greater exposure to the calmodulin (CaM) pathway correlates with longer finch beaks, suited for extracting nectar and insects, and it creates elongated chick beaks in the lab.

“We are primarily interested in the mechanistic explanations of what we see, but the interpretation is pretty exciting,” said first author Arkhat Abzhanov, assistant professor of organismic and evolutionary biology in the Harvard Faculty of Arts and Sciences, who led the project as a postdoctoral fellow in Tabin’s lab. “This allows us to say that the development of the beak is modular and tells us how it can evolve along different pathways very quickly.”

Evolution in Molecular Steps
The project started as a conversation between Tabin and Marc Kirschner, head of the HMS Systems Biology Department. Tabin is an expert on bone and limb formation who uses the chicken embryo model system. Kirschner and a co-author were wrestling with a book aimed at synthesizing developmental biology, cell biology, and evolutionary biology into a common framework now popularly known as evo/devo.

“We were asking the question, How does variation occur so selection can act?” Kirschner said. Darwin had proposed that variation occurs in random, imperceptible tiny steps. But that leaves the question seized upon by evolution opponents: How do you get complex and integrated and functional variations? Or, as Kirschner puts it, “If the name of the game is to make a beak large enough to crack a nut, and it takes a thousand variations to get there, what sustains those variations that can’t crack a nut?”

Reprinted by permission from Macmillan Publishers Ltd: Nature, Vol. 442, p. 515, 2006.

Darwin’s finches arise from a single ancestral species, but they have evolved a rich diversity of beak sizes and shapes. Two signaling molecules may control this beak development.

“Marc suggested that the work in our lab was getting to the state where we might make educated guesses about the evolution of Darwin’s finches’ beaks,” Tabin said.

When Abzhanov decided to tackle the problem, Tabin called Peter and Rosemary Grant at Princeton University. The Grants’ extensive work on natural selection on the Galapagos Islands was profiled in the Jonathan Weiner book, The Beak of the Finch. A collaboration was born. Every spring for five years, Abzhanov has accompanied the Grant teams to the restricted islands and learned how to identify birds, map the males’ mating territories, and find the nests in the sharp-thorned cactus trees.

Every day the finch lays a new egg. Of five to six eggs, laid over as many days, usually only the first two survive. On the islands, Abzhanov visited the nests of up to seven species daily, marking the first two eggs and collecting the third. He incubated the eggs, opened them, and fixed or froze them for further study back in Boston.

In one approach, he probed for genes known to be involved in craniofacial development in other species and correlated them with beak morphology. This candidate gene approach tied BMP4, one of the earliest genes expressed in beak development, to the thickness and depth of the beak. Craig Albertson of the Forsyth Institute found a similar effect of BMP4 on the jaws of cichlid fish living in small lakes in Africa. Abzhanov confirmed the BMP4 beak-building effect in chick embryo models.

High-Throughput Hunt-and-Peck
The latest paper showcases the results of the microarray approach. Using the living ancestral vampire finch for a reference and short-beaked ground finches for comparison, Abzhanov and his colleagues searched 21,000 gene transcripts for those specific to the long-billed morphology of the cactus finches. Co-author Winston Kuo, then a graduate student in Constance Cepko’s HMS genetics lab and now a postdoc in Abzhanov’s lab, wrote the software to wade through the data.

The role of many of the genes is unknown, but several signaling pathways stood out. One was represented by CaM, which senses an increase in calcium levels inside cells, activates a kinase, and launches a set of genes to make differentiation decisions. To test the pathway in chicks, the researchers made a version of the upstream CaM-dependent kinase enzyme and found that more of it increased length, but not height or depth.

Technically, all the experiments focused on the upper beaks. Yet the overexpression experiments in the chick models did not generate the monstrosities Kirschner would have expected. The bottom beak, head, neck, muscles, nerves, and other tissues seemed to go along with the change to the upper beak. “It does illustrate that these complex systems are very receptive to very simple signals,” said Kirschner, who recently published a second book on the nature of variation, The Plausibility of Life. “The tissues are all talking to each other. Here, in one step, a signal dramatically changes the whole structure of the beak, and yet it turns out to be a fairly normal-looking structure. So you don’t need as many steps of variation.”

Many interesting questions remain about the precise nature of the genetic changes responsible for the variations. But Tabin’s lab has wrapped up its work on the finches. Follow-up work is continuing in the labs of Abzhanov and another former Tabin postdoc, co-author Christine Hartmann at the Institute of Molecular Pathology in Vienna.

—Carol Cruzan Morton

top