Hens do not have teeth, and humans do not have tails. Research suggests we all have "what it takes" for a tail, and hens, indeed, have the genes that encode for teeth; however, only in very rare situations do these traits manifest themselves as a phenotype. This phenomenon is called atavism—the reappearance of a trait that had been lost during evolution. Our genes do not determine who we are, but with atavism, they can sometimes serve as reminders of our evolutionary past.
Traits that appear or disappear over time are not the result of newly mutated genes encoding defective versions of the proteins associated with teeth or tails, nor are they caused by a loss of existing genes. Instead, a growing body of experimental evidence has shown such traits reflect changes in how, where, and when these genes are expressed.
Even though birds lost teeth as physical structures between 60 and 80 million years ago, several studies have shown that those tissues within birds that would normally produce teeth still retain the potential to do so. For example, in 1821, Geoffrey St. Hilaire was the first scientist to publish the observation that some bird embryos exhibited evidence of tooth formation, but his contemporaries considered his work flawed. Since then, however, many investigators have unearthed molecular evidence that the genes involved in odontogenesis (tooth development) are indeed retained in chickens.
Despite this discovery, no one had yet demonstrated that chickens could develop teeth without external cues. This situation soon changed, however, when researchers Matthew Harris (a graduate student at the time) and John Fallon launched a study involving chickens with a particular kind of autosomal recessive mutation (Harris et al., 2006). These chickens, designated by the abbreviation ta2 for talpid-2, displayed signs reminiscent of early tooth development.
The researchers needed a positive control with which to compare their hens' teeth-that is, a closely related animal in which teeth occur. Typically, the nonmutant or "wild-type" phenotype serves as a control in gene mutation experiments, but this was an exceptional case in that the wild-type chicken doesn't have teeth. Harris and Fallon specifically needed to compare the structures they believed to be teeth in their ta2 mutant chickens with the next best thing—the closest ancestor to the chicken that still has teeth—which in this case was the archosaur, otherwise known as the common crocodile. Therefore, the researchers examined the expression of several biomarkers in wild-type chicken embryos, ta2 mutant embryos, and crocodile embryos. They found that the ta2 mutant oral cavities appeared developmentally closer to those of the crocodiles than to those of their wild-type siblings. These results thus demonstrated that all the genetic pieces to the tooth-building puzzle exist in chickens, but the directions have evolved to tell those pieces to do something different over the last 80 million years.