…in 2003, the same Japanese group published the complete sequence of the guinea pig GLO pseudogene, which is thought to have evolved independently, and compared it to that of humans [Inai et al, 2003]. 21 Surprisingly, they reported many shared mutations (deletions and substitutions) present in both humans and guinea pigs. Remember now that humans and guinea pigs are thought to have diverged at the time of the common ancestor with rodents. Therefore, a mutational difference between a guinea pig and a rat should not be shared by humans with better than random odds. But, this was not what was observed. Many mutational differences were shared by humans, including the one at position 97. According to Inai et al, this indicated some form of non-random bias that was independent of common descent or evolutionary ancestry. The probability of the same substitutions in both humans and guinea pigs occurring at the observed number of positions was calculated, by Inai et al, to be 1.84×10-12 – consistent with mutational hotspots.
What is interesting here is that the mutational hot spots found in guinea pigs and humans exactly match the mutations that set humans and primates apart from the rat (see figure below). 21,22 This particular feature has given rise to the obvious argument that Inai et al got it wrong. Reed Cartwright, a population geneticist, has noted a methodological flaw in the Inai paper:
“However, the sections quoted from Inai et al. (2003) suffer from a major methodological error; they failed to consider that substitutions could have occurred in the rat lineage after the splits from the other two. The researchers actually clustered substitutions that are specific to the rat lineage with separate substitutions shared by guinea pigs and humans. . .
If I performed the same analysis as Inai et al. (2003), I would conclude that there are ten positions where humans and guinea pigs experienced separate substitutions of the same nucleotide, otherwise known as shared, derived traits. These positions are 1, 22, 31, 58, 79, 81, 97, 100, 109, 157. However, most of these are shown to be substitutions in the rat lineage when we look at larger samples of species.
When we look at this larger data table, only one position of the ten, 81, stands out as a possible case of a shared derived trait, one position, 97, is inconclusive, and the other eight positions are more than likely shared ancestral sites. With this additional phylogenetic information, I have shown that the “hot spots” Inai et al. (2003) found are not well supported.” (see Link)
It does indeed seems like a number of the sequence differences noted by Cartwright are fairly unique to the rat – especially when one includes several other species in the comparison. However, I do have a question regarding this point. It seems to me that there simply are too many loci where the rat is the only odd sequence out in Exon X (i.e., there are seven and arguably eight of these loci). Given the published estimate on mutation rates (Drake) of about 2 x 10-10 per loci per generation, one should expect to see only 1 or 2 mutations in the 164 nucleotide exon in question (Exon X) over the course of the assumed time of some 30 Ma (million years). Therefore, the argument of the mutational differences being due to mutations in the rat lineage pre-supposes a much greater mutation rate in the rat than in the guinea pig. The same thing is true if one compares the rat with the mouse (i.e., the rat’s evident mutation rate is much higher than that of the mouse).
This is especially interesting since many of the DNA mutations are synonymous. Why should essentially neutral mutations become fixed to a much greater extent in the rat gene pool as compared to the other gene pools? Wouldn’t this significant mutation rate difference, by itself, seem to suggest a mutationally “hot” region – at least in the rat?
Beyond this, several loci differences are not exclusive to the rat/mouse gene pools and therefore suggest mutational hotspots beyond the general overall “hotness” or propensity for mutations in this particular genetic sequence.
Some have noted that although the shared mutations may be the result of hotspots, there are many more mutational differences between humans and rats/guinea pigs as compared to apes. Therefore, regardless of hotspots, humans and apes are clearly more closely related than are humans and rats/guinea pigs.
The problem with this argument is that the rate at which mutations occur is related to the average generation time. Those creatures that have a shorter generation time have a correspondingly higher mutation rate over the same absolute period of time – like 100 years. Therefore, it is only to be expected that those creatures with relatively long generation times, like humans and apes, would have fewer mutational differences relative to each other over the same period of time relative to those creatures with much shorter generation times, like rats and guinea pigs.
What is interesting about many of these mutational losses is that they often share the same mutational changes. It is at least reasonably plausible then that the GULO mutation could also be the result of a similar genetic instability that is shared by similar creatures (such as humans and the great apes).
This same sort of thing is seen to a fairly significant degree in the GULO region. Many of the same regional mutations are shared between humans and guinea pigs. Consider the following illustration yet again:
Why would both humans and guinea pigs share major deletions of exons I, V and VI as well as four stop codons if these mutations were truly random? In addition to this, a mutant group of Danish pigs have also been found to show a loss of GULO functionality. And, guess what, the key mutation in these pigs was a loss of a sizable portion of exon VIII. This loss also matches the loss of primate exon VIII. In addition, there is a frame shift in intron 8 which results in a loss of correct coding for exons 9-12. This also reflects a very similar loss in this region in primates (see Link). That’s quite a few key similarities that were clearly not the result of common ancestry for the GULO region. This seems to be very good evidence that many if not all of the mutations of the GULO region are indeed the result of similar genetic instabilities and that are prone to similar mutations – especially in similar animals.
As an aside, many other genetic mutations that result in functional losses are known to commonly affect the same genetic loci in the same or similar manner outside of common descent. For example, achondroplasia is a spontaneous mutation in humans in about 85% of the cases. In humans achondroplasia is due to mutations in the FGFR2 gene. A remarkable observation on the FGFR2 gene is that the major part of the mutations are introduced at the same two spots (755 C->G and 755-757 CGC->TCT) independent of common descent. The short legs of the Dachshund are also due to the same mutation(s). The same allelic mutation has occurred in sheep as well.