A bat’s elongated paw results from altered interactions of ancient genes. Illustration: Alison Schroeer
Consider the relationship between bat and mouse. Separated from their last common ancestor — a little mouse-like thing — by somewhere between 80 million and 100 million years, the two have had plenty of time to pick up new habits. Bats, beautifully adapted fliers, flit through the sky on membranous wings stretched over long limbs and outsized paws. Mice, those small, earthbound scurriers, scuttle about on short, swift legs. Although the two animals have forelimbs that differ in obvious ways, closer inspection reveals profound homologies: The same bones are used to build both wing and leg. Both have a humerus, a radius, and an ulna, and a collection of carpus, metacarpus, and phalanges, all in the same relationship to one another and having recognizable affinities, and all developing under the guidance of the same set of genes.
In the millions of years since the bat diverged from the mouse, evolution has flattened, stretched and drawn out the mouse-like paw to form a fan. We biologists don’t know the precise steps by which the bat’s wing evolved, but recent work comparing the molecules that regulate the growth of the limb in mice and in bats has begun to reveal how small changes led from one form to the other. At the same time, the research has lent strong support to some of evolutionary biology’s leading hypotheses for how evolution and development interact, and for how genotypes and phenotypes can evolve at such apparently disparate rates.
One small piece of the story is a gene called Prx1. Prx1 plays a variety of roles throughout a developing embryo but has a particularly important role in growing bones; it becomes active in limbs shortly after limb buds begin to form, at the same time that bones begin to condense and elongate. Mutant mice that lack Prx1 suffer from, among other defects, grossly shortened limbs. Prx1 is also present in the bat, and it’s expressed in the same places but with some significant differences. In bats it is active over a longer stretch of the forelimb, and it’s also turned on much more strongly in the hand, coincident with the remarkable elongation of the digits into supports for the wing. These observations all suggest that differences in Prx1 expression may contribute to the differences in forelimb shape between a bat and a mouse.
Now for one little problem: The bat Prx1 and mouse Prx1 genes are almost identical! There are only two amino-acid differences between the two proteins, and neither of those alterations is located in regions of the protein that have significant binding activity. This suggests that the two genes do exactly the same thing, but that they are regulated differently — that there is variation in the parts of the genome that turn mouse and bat Prx1 on and off. It’s key, then, to look at regions of DNA in the neighborhood of the protein-coding portion of Prx1 and find the switches that control it. That is where we expect to see a difference.
Chris Cretekos, of Idaho State University, and others went searching through the regions around the Prx1 gene in the mouse (Mus musculus) and the short-tailed fruit bat (Carollia perspicillata) and found a short stretch of DNA, not a part of the protein-coding region, that was largely conserved in bat and mouse, suggesting an important function was constraining evolution. Nevertheless the region, called an enhancer, was not identical in the two species, which indicates some potential for a significant evolutionary and developmental change — and that it might be responsible for some of the differences in bats and mice.
To confirm this Cretekos and his colleagues carried out a brilliant experiment. They removed the Prx1 enhancer region from a mouse and replaced it with the Prx1 enhancer from a bat. Basically, they switched the switches, putting a bat regulator into a mouse, and then let the system develop a limb.
Before I tell you what the result was, let me tell you what the result was not: They did not get baby mice with bat wings. Which wasn’t what they expected. Prx1 is one gene of many that contribute to limb formation, and its enhancer is only a small section of a larger control region, so Cretekos and his colleagues anticipated an incremental shift in limb proportions. And that’s what they got: The limbs of the embryonic mice with the bat enhancer were 6 percent longer than those of mice with the normal mouse enhancer. Furthermore, the length increase was temporary; as the mice grew into adulthood, the limbs grew into normal mouse-like proportions. In an additional surprise, deleting the mouse Prx1 enhancer altogether produced mice with normal limbs as well.
Such a small effect might not seem particularly significant (or interesting!), but this is a beautiful result, fitting perfectly into a number of ideas current in evo-devo and in conventional evolutionary biology as well.
For one thing, this experiment highlights the importance of regulatory changes in evolution. No changes were made to the mouse gene itself; the only changes were to regulatory switches for the gene, which, because the bat and mouse forms of Prx1 are so similar, seems to be how evolution built a wing from a paw: Changes in gene expression alone altered the development of the limb. (Nevertheless changes to the coding regions of genes are still important in other aspects of evolution.)
The experiment further confirmed that structures in organisms, such as wings or legs, are not the product of simple, single genes, but of many interacting genes that produce a complete organ. This should be self-evident, but simplistic versions of genetics still tend to talk about genes “for” structures and behaviors, such as eyes or schizophrenia. Prx1 seems to be a gene that regulates cell division in many tissues, including growing bones, and it is only one among many such regulators; slight changes in Prx1 expression will be diffuse and partial.
Cretekos also confirmed the important role of functional redundancy — that is, some changes may have no observable developmental effect, simply because other genes in the system can fill in and take up any slack. Individual genes, too, can have multiple regulatory elements with overlapping functions. Deleting the mouse Prx1 enhancer, for instance, had no detectable effect because other regulatory elements still permitted normal expression of the gene. Furthermore, development is canalized: Redundant developmental mechanisms buffer the system against change. Single changes in the genome may produce small, temporary changes in the phenotype, but the totality of the processes will converge on a consistent, predictable result.
What this suggests about evolution is a combination of the gradual and the rapid. Mutations can accumulate slowly and imperceptibly in a population in a gradual way, but that morphological change can occur relatively rapidly by the recombination of forms of genes that alone do little but together cause a swift transformation. Developmental and evolutionary innovation arises from the gathering of incremental predispositions in a diverse population that can, under selection, be shuffled into advantageous configurations, but the small steps of simpler genetic changes are what we expect to see in the laboratory and the field.
Originally published August 27, 2008