Synteny refers to how well chromosomal sequences from different species align with one another. Genes can be in the same general order and location between species, for example rat and mouse, or chimp and human. If they align well evolutionists take the alignment as evidence for common ancestry. Sometimes, the gene sequences may be interrupted by deletions or insertions, and stop codons, which prevent the gene from making functional protein. These “inactivated” genes are called pseudogenes, and are taken by evolutionists as further evidence for common descent. Their presence is explained as the remnants of once functional genes broken by mutation and no longer needed by the organism.
“Why should a designer litter the genome with so much trash?” the evolutionists say.
One particular story that combines ancient pseudogenes and synteny to argue for common descent is circulating on the web, right now. The story is several years old, but it’s popular because it seems to confirm the old evolutionary idea that our long ago ancestors were egg-laying creatures that then evolved to live-bearing placental mammals. And it’s being challenged because of new evidence coming from ENCODE and other genome projects.
A blog by Dennis Venema lays out the supposed story: humans retain the remnants of DNA that used to code for egg yolk proteins called vitellogenins. Since humans don’t lay eggs, the argument goes, these “vitellogenin” pseudogenes, now long since mutated into near unrecognizability, must be inherited from common ancestors who did lay eggs. Venema claims that all three human “vitellogenin” pseudogenes, VIT1 through VIT3, show traces of sequence similarity to the functioning vitellogenin genes of the chicken.
All of this argument depends on three ideas: 1) we share common ancestry with egg laying animals, 2) any sequence similarity is due to ancestry not function, and 3) there was just one function for the vitellogenin gene, and that was making vitellogenin. Any function now is due to secondarily evolved function.
This story is based on a 2008 paper by Brawand et al. that discusses yolk proteins in egg-laying animals and mammals. In that paper they identified the region of the chicken genome where vitellogenin genes are located, then found the similar regions in humans, dogs and various marsupials and platypus, to see if vitellogenin pseudogenes could be found in the right syntenic neighborhoods.
Patches of sequence similarity to the chicken genome that might be interpreted as pseudogenes can be found in syntenic regions of marsupial genomes. Evidence for vitellogenin pseudogenes in human and dog genomes, on the other hand, is very weak. A graph from the paper shows where the human (Y axis) and chicken (X axis) DNA sequences align. The top panel illustrates the putative VIT 1 alignment, and the bottom V2 and 3. Black dots show where some sequence similarity exists. The colored stripes indicate coding sequences (exons) for each “gene.” In between in the white regions are the non-coding introns. ELDT1, FP, SSX2IP and CTBS are functional genes in both human and chicken and show sequence similarity in all coding sequences.
Synteny is practically nonexistent for vitellogenin genes VIT2 and VIT3 (it is not statistically significant compared to the genomic background). The remaining gene, VIT1, has two patches of similarity in its putative coding sequence, at either end of the”pseudogene” sequence. The best of them is a patch 150 bases long (out of 42,637 total bases for the gene!) that has roughly 50% identity, by my estimation, and a few deletions to help make things match up. According to the authors, there is a 5% chance that the amount of similarity between VIT1 for humans and dogs, and the chicken VIT1, is due to random chance–that’s just at the borderline for statistical significance.
What apparent similarity there is could well be due to an overlapping gene with an entirely different function that is present in that stretch of sequence in the chicken and human genomes. Indeed, there is evidence of another gene with other possible functions in that region of the genome, including the 150 base stretch mentioned above. As reported by Tomkins,
…the alleged vtg [vitellogenin] fragment in human is not a pseudogene remnant at all, but a functional enhancer element in the fifth intron of a “genomic address messenger” (GAM) gene. This GAM gene produces long noncoding RNAs that have been experimentally shown to selectively inhibit the translation of known target genes, a majority of which are implicated in a variety of human diseases. Messenger RNAs from this gene are also expressed in a variety of human brain tissues. The alleged 150 base vtg sequence contains a variety of highly conserved mammalian transcription factor binding domains, nucleosome depleted open-active chromatin, is hypo-methylated, associates with RNA polymerase 2 in long-range chromatin interactions, and binds the Mafk transcriptional regulator. These combinatorial data clearly show that it is a functional enhancer element in a GAM gene expressed in the human brain–strongly challenging the idea that this sequence is an egg-laying pseudogene genomic fossil.
The long non-coding RNAs mentioned above are believed by many to have many important regulatory functions in the cell. They are implicated in long and short range interactions between genes, the way the DNA loops, whether genes are sequestered or not, all these things and more are affected.
Let’s summarize some of the main points we have discussed.
1. A similar arrangement of genes exists among mammalian genomes and the chicken genome in the neighborhood of the chicken vitellogenin genes, with the very important exception that in mammals, the vitellogenin genes are either absent or pseudogenized.
2. This similar order could be due either to ancestry or functional reasons.
3. In humans the evidence for a vitellogenin pseudogene fragment is very weak.
4. It is possible that any sequence similarity there might be due to a shared function, because of the regulatory gene that exists in the same region. (This despite Venema’s disparaging comments to the contrary.)
5. If the sequence similarity is due to function, then it need not indicate common ancestry, or the former presence of an egg-laying protein in the human genome.
6. There is insufficient sequence similarity to make the claim for a vitellogenin pseudogene.
7. Claiming exaptation, a la Venema, is a post hoc argument based on the prior assumption of common descent. His argument goes like this:
There is sequence similarity to vitellogenins in the human genome.
It must be due to common descent.
There is evidence of another possible function for the sequence.
It must have evolved as a secondary function, because we know that it was originally vitellogenin.
Claiming a functional reason for the similarity is a Creationist dodge.
My response: Claiming exaptation sounds like a dodge to me.
Bottom line: the vitellogenin pseudogene is pseudo, and shouldn’t be used as evidence for common descent. So the next time you see this story trotted out as proof that we evolved from an egg-laying ancestor, feel free to say, “Big claim for such weak evidence!”