As we saw in yesterday’s post, the 2005 Kitzmiller case was a watershed for the ID movement. Central to the case was demonstrating to Judge Jones that evolution, in contrast to ID, was a productive scientific theory. Part of that case was made by biologist Ken Miller, who showed the judge a few lines of evidence for shared ancestry of humans and great apes. These examples were intended to show how a relatively new science (genomics) could be used to test evolutionary predictions based on other lines of evidence (such as anatomy and physiology). One of Miller’s examples, as we saw yesterday, was the beta globin η pseudogene shared between humans and chimpanzees.
In 2013, further research revealed that this pseudogene may in fact have a function. Casey Luskin of the Discovery Institute was quick to claim that these results rendered Miller’s argument null and void, a claim he has now repeated on the 10th anniversary of the Kitzmiller decision:
Ken Miller's argument for Darwinian evolution, and against ID, depends on the beta-globin pseudogene being "non-functional," implying as he does that the genetic differences between it and protein-coding beta-globin genes are errors. In light of this new evidence for the functionality of the beta-globin pseudogene, it seems that those genetic differences may not be errors at all. If so, then Miller's argument, his Exhibit A, collapses.
Though Luskin’s argument is sparse, we can glean the essential points: if the beta globin η pseudogene has a function, then we infer that we see this sequence in the human genome because it was deliberately designed to be there in its observed form. It would then come as no surprise that this functional gene is present in chimpanzees as well, since they too need functional beta globin genes. Since Miller’s original argument based on this evidence was for shared ancestry for humans and chimpanzees, it seems that that Luskin views his rebuttal as supporting independent design (i.e. special creation) of humans and chimpanzees, though he does not spell this out explicitly.
The paper (PDF) that Luskin cites is an interesting study that set out to look for sequences in the human beta globin cluster that show (a) less variation among humans than we would expect than if the sequences were of no functional importance; and (b) show conservation with the chimpanzee genome (i.e. sequences that have been selected for since our two lineages went their separate ways about 4-6 million years ago). Not surprisingly, the functional beta globin genes are constrained by both measures – they show far less variation than one would expect if the sequences were free to mutate without consequence to the organism. What was somewhat surprising to the authors of the study was the finding that this sequence conservation extends to the beta globin η pseudogene as well. Obviously, the function of this sequence cannot be to make a beta globin protein, since it has many mutations in its coding sequence that would prevent that. As an alternate hypothesis, the authors of the study look for evidence that this sequence is acting as regulatory DNA: DNA that helps control how and when the functional beta globin genes are used to make proteins. They find some good evidence to support this hypothesis. As such, they conclude that the beta globin η pseudogene sequence functions as regulatory DNA in humans (and likely in chimpanzees and other primates as well).
This is the evidence that Luskin uses to claim Miller’s argument for common ancestry in Kitzmiller is flawed. So, does Luskin’s argument stand up to scrutiny?
The first thing that strikes me as odd about Luskin’s argument is this: even if the beta globin pseudogene has a function different from the very similar, transcribed and translated beta globins genes surrounding it, it seems a pretty big stretch to claim that this sequence was designed independently of it. There are tens of thousands of genes in the human and chimpanzee genomes; all of them require regulatory DNA sequences for their proper expression. Yet only this gene, in the entire human (or chimpanzee) genome, has regulatory DNA that is similar to a translated beta globin gene. Why is it that the designer used this regulatory sequence for this gene, and only this gene, when there are tens of thousands of other ways to construct a regulatory DNA sequence?
Perhaps, the argument might go, the human and chimpanzee beta globin gene clusters need regulatory DNA that just happens to look like a beta globin coding sequence in order to function, because beta globin clusters in general require such sequences. That’s an interesting hypothesis, and one that is testable. A simple perusal of genome databases, however, shows that many other mammals have beta globin gene clusters like humans and chimpanzees do, except they do not have any sequences like the beta globin η pseudogene. So, from an ID perspective, one would conclude it’s easy to design a fully functional beta globin gene cluster without using regulatory DNA that happens to look like a degraded beta globin coding sequence. Such designs are found in many other mammals.
Perusing genome databases further, we would notice something else: some mammals, such as goats, do have sequences very like the human beta globin η pseudogene, and in the same location in their beta globin cluster – but in their case, the sequence is expressed as a functional beta globin protein! So the designer seems to have used this sequence in some mammals to form a protein, in other mammals in a modified way to form regulatory DNA, and in other mammals not at all. A diagram can help us see these different design states (redrawn from Figure 4 in this paper):
If all this seems a little strained, you’re not alone in thinking so. Of course, a design argument can be used to explain any pattern we see in nature, if for no other reason that the designer, for reasons unknown to us, wanted it that way. Such an explanation, however, often ends up becoming very ad hoc.
Of course, there is another way to look at the data – that the “designer” – in other words, from a Christian perspective, God – was pleased to make the various mammalian beta globin gene clusters through the process of evolution. In an evolutionary scenario, things make a lot more sense, as a subtle adjustment to our diagram will help us see:
The reason we see a beta globin η sequence in the human and chimpanzee genomes (as well as in the genomes of all other primates sequenced thus far) is because this sequence was once transcribed and translated as a beta globin protein, as we expect for a sequence so similar to a coding sequence. This is a far less strained interpretation than suggesting this regulatory DNA, though unlike any other regulatory DNA in the human genome, just happens to sit among coding sequences it resembles, even though thousands of other regulatory sequences would work equally well. We see beta globin η as a functional gene in goats because in their lineage this gene was not mutated and lost. This suggests that the last common ancestor of humans and goats had a functional beta globin η gene – which further suggests that this gene was lost entirely in the rabbit lineage (and underscores the point that using a degraded beta globin η sequence is not the only way to make functional regulatory DNA in this region).
What Luskin fails to consider in his argument is exaptation – the conversion of a structure from one function to another through mutation and selection. In this case, we have good evidence that the beta globin η pseudogene in primates has been exapted from a coding gene into regulatory DNA.
As such, the “shared mistakes” that Miller (and Behe) refer to are still properly understood as mutations that inactivated the protein coding function of the beta globin η gene in the primate lineage prior to the divergence of the lineages leading to humans and chimpanzees. As such, these mutations remain evidence for common ancestry, and neither Miller nor Behe need to change their argument (and I note that Behe has not, to my knowledge, done so despite Luskin’s repeated use of this example). So, not even leading ID figures seem to find Luskin’s case compelling.
The general strategy of proclaiming any found function for a pseudogene as evidence against common ancestry is a common one in antievolutionary apologetics, not merely within the ID movement. In the coming weeks, we’ll explore another example in depth – the recent claim by a young-earth creationist that the human vitellogenin pseudogene is in fact not a pseudogene at all.
For further reading (two series)