So you don’t think evolution is controversial?
For those who are bound and determined to view evolution as scandalous, here is a perfectly legitimate controversy within the discipline that just may be swelling up into a full-scale storm. The issue at hand is whether mutations in cis-regulatory regions (DNA sequences near genes that regulate their expression) or mutations in coding regions (DNA sequences that encode functional molecules) have been more important in generating the phenotypic variation necessary for diversification and adaptation. Of all the different types of DNA that make up genomes, we understand coding DNA the best. So not surprisingly, many of the mutations discovered to produce differences in phenotype—like those known to cause coat color variation in domestic cats and jaguars—have been found in coding DNA.
However, some examples of phenotypic variation map to changes in DNA surrounding the gene of interest, which control not what protein the gene encodes but how much of the protein is produced. A notable example is the bony armor in threespine stickleback fish: the difference between sticklebacks with pronounced armor and those with none has been traced to a cis-regulatory region for the gene Pitx1, a transcription factor involved in skeleton formation. Some “evo-devo” researchers who investigate evolutionary questions using these developmental biology techniques argue that striking changes in morphology, like those in threespine sticklebacks, are, in principle, more likely to be due to cis-regulatory mutations than protein-coding mutations. And thus it follows that cis-regulatory mutations must have been more evolutionarily important than protein-coding mutations in the diversification and adaptation of organismal forms.
The argument for cis-regulatory evolution is not new; this idea has garnered attention and enthusiasm since, for example, Sean Carroll published an article on the evolution of gene regulation in 2000. Criticism for this enthusiasm isn’t new either; in 2005, Jerry Coyne wrote a critical review of another of Carroll’s books, which itself was reported on in this blog. And of course myriad other articles, responses and critiques have filled the peer-reviewed and popular science literature. But the latest effort comes again from Coyne and co-author Hopi Hoekstra, in an article published in Evolution this month. In it, Hoekstra and Coyne argue that while mutation in cis-regulatory regions may be important, we don’t have nearly enough evidence to declare it the predominant vehicle for macroevolutionary change. What do you think?
May 8th, 2007 at 5:18 pm
I guess the question is, do we really have to choose a “predominant vehicle”? Phenotypic variation is fundamentally a gradient determination, right? I suspect that we’ll discover that the salient ‘traits’ we identify in some species will be products of coding DNA and some will be the products of regulatory regions. In the case of the stickleback, for example, the difference in the armor is one of “pronouncement”…a scalar notion. Shouldn’t we expect traits with a more mendelian distribution to be ‘coded’ whereas gradient traits are ‘regulated’?
May 9th, 2007 at 11:51 am
It’s an interesting point, but I’m not sure that it will prove to be true. For example, discrete traits with Mendelian inheritance (that is to say, non-quantitative traits) may still be due to changes in regulatory regions–I think many of the antibiotic resistance traits in bacteria may be an example of this. I certainly do agree that at the end of the day, both regulatory and protein-coding mutations will have been shown to be important in evolution. The question then just becomes an issue of proportion–and as one wise biologist in my department just pointed out, this is a more boring question.
May 10th, 2007 at 2:26 pm
I think the most encouraging aspect of this debate is that it has exposed the weakness of the “reverse genetics” approach to evolution that has been the modus operandi for the last few decades — i.e., pick your favorite polymorphic trait, and then figure out the mutations that caused it. Whether or not this approach was biased in favor of a certain type of trait (form vs. function, discrete vs. graded, or whatever), published results certainly favored the causative mutations that were easiest to find, more often than not in protein-coding regions. When Carroll, Stern, et al. started looking at cis-regulatory regions a few years, they may have oversold their case as a general model for evolution simply to overcome this historical bias. Now that we have all these genomes being sequenced, we actually have the power to resolve this debate using an unbiased, “forward genetics” approach by looking for evolutionary changes at the sequence level with no a priori classification of biological function. That’s really exciting — and if any patterns emerge, one side of the debate can have some icing on their cake.
May 11th, 2007 at 5:07 pm
[...] it looks like the anti-evolutionists at Uncommon Descent have jumped on the regulatory/coding DNA controversy. Well… no not really. Following the publication of an article in New Scientist that describes [...]
July 1st, 2011 at 9:39 am
[...] or protein-coding regions in DNA have been more important in the evolution of morphology. (Click here for a snippet of this discussion.) These internal conflicts reflect a gap in our knowledge [...]