They selected teosinte...and got corn. Excellent!

Evolutionary science is so much bigger, so much deeper, so much more interesting than its opponents (understandably) will admit. It's more complicated than Michael Behe or Bill Dembski let on, and yet it's not that hard to follow, for those who are willing to try. The best papers by evolutionary biologists are endlessly fascinating and scientifically superb, and reading them is stimulating and fun.

Yet, as an experimental developmental biologist reading work in evolutionary biology, I often find myself yearning for what we call "the definitive experiment." Molecular biology, for example, can point to a few definitive experiments -- elegant and often simple -- that provided answers to big questions. Sometimes, while examining an excellent evolutionary explanation, I think, "Wouldn't it be great if they could do the experiment?"

Now of course, plenty of evolutionary biology is experimental, and I've reviewed some very good examples of experimental evolutionary science on this blog. But when it comes to selection and the evolution of new structures and functions, the analysis often seems to beg for an experiment, one that is simple to conceive but, typically, impossible to actually pull off -- there's not enough time. The previous Journal Club looked at one way around this limitation: bring the past back to life. Even better, though, would be to find an example of evolutionary change in which the new and old forms are still living, so that one could do the before-and-after comparison. It would look something like this: take a species, subject it to evolutionary influences of some kind until the descendants look significantly different from the ancestors, then compare the genomes (or developmental processes) of the descendant and the ancestor, in hopes of discovering the types of changes at the genetic or developmental level that gave rise to the differences in appearance or function of the organisms. That would be a cool experiment.

In fact, that kind of experiment has been done, more than once. The best example, in my opinion, involves an organism far less sexy than a dinosaur or a finch or a whale: Zea mays, better known as corn (or maize).

Corn is a grass, but a grass that's been so extensively modified genetically that it's barely recognizable (to non-specialists like me) as a member of that family. Wait...genetically modified? Yes, and I'm not talking about the really modern tricks that gave us Bt corn or Roundup Ready corn. In fact, the wonderful stuff they grow in Iowa is quite different from the plants that humans first started to harvest and domesticate in Central America a few millenia ago. Corn as we know it is the result of a major evolutionary transformation, driven by selection at the hands of humans. (I don't find the natural/artificial selection distinction at all useful, since there's no explanatory difference, but you can refer to the selection under consideration here as 'artificial' if it makes you feel better.) The story has been a major topic in evolutionary genetics for decades, but it's largely absent from popular discussions, probably because the Discovery Institute has wisely avoided it. I hope it will soon be clear why you won't find the word 'teosinte' anywhere at discovery.org.

For many years, the origin of corn was a mystery. Like most known crops, it was domesticated 6000-10,000 years ago. But unlike other crops, its wild ancestor was unknown until relatively recently. Why this odd gap in our knowledge? Well, it turns out that corn is shockingly different -- in form, or morphology -- from its closest wild relative, which is a grass called teosinte, still native to southwestern Mexico. In fact, corn and teosinte are so different in appearance that biologists initially considered teosinte to be more closely related to rice than to corn, and even when evidence began to suggest a genetic and evolutionary relationship, the idea was hard to accept. As John Doebley, University of Wisconsin geneticist and expert on corn genetics and evolution, puts it: "The stunning morphological differences between the ears of maize and teosinte seemed to exclude the possibility that teosinte could be the progenitor of maize." (From 2004 Annual Review article, available on the lab website and cited below.)

But it is now clear that teosinte (Balsas teosinte, to be specific) is the direct ancestor of corn. In addition to archaeological evidence, consider:

• The chromosomes of corn and teosinte are nearly indistinguishable at very fine levels of structural detail.
• Analysis using microsatellite DNA (repetitive DNA elements found in most genomes) identified teosinte as the immediate ancestor of corn, and indicated that the divergence occurred 9000 years ago, in agreement with archaeological findings.
• Most importantly, a cross between corn and teosinte yields healthy, fertile offspring. So, amazingly, despite being so different in appearance that biologists initially considered them unrelated, corn and teosinte are clearly members of the same species.

The basic idea, then, is that corn is a domesticated form of teosinte, exhibiting a strikingly distinct form as a result of selection by human farmers. And that means that we have a perfect opportunity to examine the genetic and developmental changes that underlie these "stunning morphological differences." We can do the experiment.

First, have a look at an example of one of the evolutionary changes in teosinte under human selection.

The small ear of corn on the left is a "primitive" ear; the brown thing on the right is an ear from pure teosinte. (Both are about 5 cm long.) The "primitive" ear is similar to archaeological specimens representing the earliest known corn. Image from John Doebley, "The genetics of maize evolution," Annual Review of Genetics 38:37-59, 2004. Article downloaded from Doebley lab website.

The thing on the far left is a teosinte "ear," the far right is our friend corn, and the middle is what you get in a hybrid between the two. Photo by John Doebley; image from Doebley lab website.

The pattern of branching of the overall plant is also strikingly different between corn and teosinte, and you can read much more on the Doebley lab website and in their publications.

When I first heard about this work at the 2006 Annual Meeting of the Society for Developmental Biology, I was astonished at the amount of basic evolutionary biology that was exposed to experimental analysis in this great ongoing experiment. Here are two key examples of the insights and discoveries generated in recent studies of corn evolution.

1. Does the evolution of new features require new, rare, mutations in major genes?

Perhaps this seems like a stupid question to you. Anti-evolution propagandists are eager to create the impression that evolutionary change only occurs when small numbers of wildly improbable mutations somehow manage to help and not hurt a species. And in fact, experimental biology has produced good examples of just such phenomena. But there is at least one other genetic model that has been put forth to explain the evolution of new forms. This view postulates that many major features exhibited by organisms are "threshold" traits, meaning that they are determined by many converging influences which add together and -- once the level of influence exceeds a threshold -- generate the trait. The model predicts that certain invariant (i.e., never-changing) traits would nevertheless exhibit significant genetic variation, since evolutionary selection is acting on the overall trait and not on the individual genetic influences that are added together. Hence the implication that...

...populations contain substantial cryptic genetic variation, which, if reconfigured, could produce a discrete shift in morphology and thereby a novel phenotype. Thus, evolution would not be dependent on rare mutations, but on standing, albeit cryptic, genetic variation.

--from Nick Lauter and John Doebley, "Genetic Variation for Phenotypically Invariant Traits Detected in Teosinte: Implications for the Evolution of Novel Forms," Genetics 160:333-342, 2002.

In that paper, the authors show that several invariant traits (e.g., number of branches at the flower) in teosinte display significant genetic variation. In other words, the traits are the same in every plant, but the genes that generate the traits vary. The variation is 'cryptic' because it's not apparent in basic genetic crosses. But it's there. The authors ask: "How can cryptic genetic variation such as we have detected in teosinte contribute to the evolution of discrete traits?" Two ways: 1) the variation is available to modify or stabilize the effects of large-effect mutations; and 2) variation in multiple genes can be reconfigured such that it adds up to a new threshold effect. Note that the first scenario is clearly applicable to the kind of evolutionary trajectory outlined by Joe Thornton's group and discussed in a previous post. The second scenario is particularly interesting, however, since it addresses an important question about the role of selection. Consider the authors' discussion of this issue:

At first glance, cryptic variation would seem inaccessible to the force of selection since it has no effect on the phenotype. However, if discrete traits are threshold traits, then one can imagine ... that variation ... could be reconfigured such that an individual or population would rise above the threshold and thereby switch the trajectory of development so that a discrete adult phenotype is produced. We find this an attractive model since evolution would not be constrained to “wait” for new major mutations to arise in populations. (Italics are mine; ellipses denote deletion of technical jargon, with apologies to the authors.)

In fact, in a 2004 review article, Doebley is bluntly critical of the assumption that new mutations were required during the evolution of corn, and seems to suggest that this view led researchers significantly astray:

There is an underlying assumption in much of the literature on maize evolution that new mutations were central to the morphological evolution of maize. The word "mutation" is used repeatedly to describe the gene changes involved, and Beadle led an expedition ("mutation hunt") to find these rare alleles. The opposing view, that naturally occurring standing variation in teosinte populations could provide sufficient raw material for maize evolution, was stated clearly for the first time by Iltis in 1983. Although new mutation is likely to have made a contribution, anyone who has worked with teosinte would agree that teosinte populations possess abundant genetic variation. [...] Allowing for cryptic variants and novel phenotypes from new epistatic combinations to arise during domestication, it is easy to imagine that maize was domesticated from teosinte.

--John Doebley, "The genetics of maize evolution." Annual Review of Genetics 38:37-59, 2004.

Compare that discussion, and others like it in the paper I'm quoting, with the yapping about mutations that passes for anti-evolution criticism of evolutionary genetics. I can find no evidence that Michael Behe or any other ID theorist has even attempted to seriously address the importance of genetic variation in populations. I haven't read The Edge of Evolution yet, but I have it right here, and the index suggests that Behe hasn't tried to engage genetics beyond the high school level. There's a good reason why Behe is an object of scorn in evolutionary biology. He wants you to think it's because his critics are mean. No; it's much worse than that.

2. Does evolutionary change ever result from a "gain of information," or does Darwinian evolution merely prune things out?

It would be easy to get the impression from various creationists and ID proponents that mutation and selection can only remove things from a genome. Young-earth creationist commentary on "microevolution" (a yucky term for the now-undeniable fact of genetic change over time) always adds that this kind of change involves NO NEW INFORMATION. (The caps are important, apparently, since caps and/or italics are de rigueur in creationist denialism on this topic.)

Similarly, Michael Behe wants you to think that beneficial (or adaptive) mutations are some kind of near impossibility, and that when they do happen it's almost always because something's been deleted or damaged, with a beneficial outcome.

Studies of evolution in corn and teosinte (and other domesticated plants), not to mention findings like the HIV story on Abbie Smith's now-famous blog, tell a different -- and, of course, more wonderfully interesting -- story. In a minireview on the genetics of crop plant evolution in Science last June, John Doebley notes that most of the mutations that led to major evolutionary innovations occurred in transcription factors, which are proteins that turn other genes on and off. Then this:

Another remarkable feature of this list is that the domesticated alleles of all six genes are functional. If domestication involved the crippling of precisely tuned wild species, one might have expected domestication genes to have null or loss-of-function alleles. Rather, domestication has involved a mix of changes in protein function and gene expression.

In other words, the new genes are not dead or damaged; they're genes that are making proteins with new functions. ('Allele' is just the term for a particular version of a particular gene, and 'null', as you might have guessed, is a version that is utterly functionless, as though the gene were deleted entirely.) Now, if you've even flipped through The Origin of Species, you might not be surprised by Doebley's conclusion:

Given that the cultivated allele of not one of these six domestication genes is a null, a more appropriate model than "crippling" seems to be adaptation to a novel ecological niche -- the cultivated field. Tinkering and not disassembling is the order of the day in domestication as in natural evolution, and Darwin's use of domestication as a proxy for evolution under natural selection was, not surprisingly, right on the mark.

The change from teosinte to corn happened in about a thousand years. That's fast evolution. Apply selection to a varying population, and you get new functions, new proteins, new genes, completely new organisms. Fast.

So in summary, we can do the experiment. And we've done the experiment. ('We' being John Doebley and his many able colleagues.) And we've learned a lot about evolution and development. Now if we can just get people to read it. Then they'll know more about evolution, and about God's world, and about the trustworthiness of the anti-evolution propaganda machines that are exploiting the credulity of evangelical Christians.