Pseudogenes

A pseudogene is a sequence that retains the recognisable structure of a working gene but has been disabled — by a premature stop codon, a frameshift, a lost promoter, or accumulated substitutions that leave the reading frame intact but the product useless.

Genomes carry large numbers of them. The interesting cases are the ones where the gene still works in related animals, so its former job is known, and where the specific disabling mutation can be compared across species.

Shared function proves little on its own; two designers solving the same problem might reach the same solution. Shared damage at the same nucleotide is a different kind of evidence, and it is the reason this line of argument is difficult to answer.

Vitamin C, lost repeatedly#

Most mammals synthesise their own vitamin C. The final step is catalysed by L-gulonolactone oxidase, encoded by GULO.

The pattern is the argument. Ohta and Nishikimi (1999) compared primate GULO remnants and found the same deletions and substitutions in the same positions across the group — one breakage inherited by all of them. Guinea pigs are also scurvy-prone, but their copy is broken differently, exactly as an independent loss should look.

Bats make the same point with better resolution. Cui et al. (2011) found that GULO has been lost independently in several bat lineages, at different times and by different mutations, while others retain a working copy.

Losses are not distributed randomly across the tree, and identical losses are not scattered across unrelated groups. Each break sits on one branch and is inherited by everything below it.

Birds that carry the machinery for teeth#

No living bird has teeth. Birds have lacked them for roughly 100 million years, and early birds such as Archaeopteryx had them.

The genes are still there. Meredith et al. (2014) sequenced tooth-specific genes across 48 bird species and found the enamel and dentine genes present as pseudogenes in all of them, with shared inactivating mutations indicating a single loss in the common ancestor of living birds — a molecular date consistent with the fossil record.

The developmental counterpart is documented in Vestigial Structures and Atavisms: the naturally occurring talpid² chicken mutant forms conical tooth buds, and chicken tissue can be induced to begin forming tooth structures experimentally.

Whales#

Baleen whales filter-feed and have no teeth as adults, though their embryos develop tooth buds that are later resorbed.

Meredith et al. (2011) found that the enamel gene MMP20 is a pseudogene in every baleen whale examined, disabled by mutations shared across the group. Deméré et al. (2008) combined that molecular evidence with fossils of Aetiocetus, an extinct whale that had both teeth and the nutrient foramina associated with baleen.

Cetacean olfactory receptor genes tell a parallel story. Hayden et al. (2010) surveyed olfactory subgenomes across mammals and found the fraction of pseudogenes tracking ecology — highest in animals that rely least on smell — with the specific broken copies shared according to relatedness rather than lifestyle.

Egg yolk genes in mammals#

Placental mammals do not lay eggs and produce no yolk. Vitellogenin, the egg-yolk protein, is central to egg-laying vertebrates.

Brawand, Wahli and Kaessmann (2008) found vitellogenin pseudogenes in the human, dog, and other placental genomes, in the chromosomal locations that synteny with the chicken genome predicted — sequences that were looked for because their position could be predicted in advance. The platypus, which still lays eggs, retains one functional copy. The paper ties the loss to the origin of lactation and placentation.

Taste receptors in penguins#

Zhao, Li and Zhang (2015) found that all penguins have lost sweet, umami, and bitter taste reception, retaining only sour and salty. The receptor genes are present as pseudogenes, with shared lesions indicating loss in the common ancestor of the group.

Loss of umami reception in an animal that eats nothing but fish is difficult to read as a design decision. As an inherited accident in a cold-water lineage where the relevant receptor proteins function poorly, it is unremarkable.

Common objections#

"Pseudogenes have functions — 'junk DNA' was a failed prediction"

Some sequences classified as pseudogenes are transcribed and a number have regulatory roles. This is established and not disputed, and the ENCODE project's 2012 report that most of the genome is "biochemically active" is often cited here.

ENCODE's definition of activity covered any reproducible transcription or protein binding, which occurs at low levels across much of the genome. Graur et al. (2013) criticised the equation of that with function in detail.

More importantly, the objection does not reach the argument. For GULO, avian enamel genes, and whale MMP20, the original function is known and demonstrably absent: primates get scurvy, birds have no teeth, baleen whales have no enamel. Whatever secondary roles these sequences acquired, they no longer do the job their structure is built for — and the argument rests on the distribution of the specific disabling mutations, which function does not address.

"A designer could have disabled genes deliberately"

Possible, and unfalsifiable as stated.

It does not predict the pattern. Deliberate disabling would be expected to track need — animals that do not require a gene would lack it, and the manner of disabling would be arbitrary.

What is observed is that the manner of disabling is shared by relatedness. Every haplorhine primate carries the same GULO lesions; guinea pigs, which have the same dietary consequence, carry different ones. A design account has to explain why the broken copies sort into a tree rather than by function.

"Loss of function is degeneration, not evolution"

Correct as a description. These are losses, and creationist authors readily accept that genomes accumulate damage.

Both accounts predict damage. They differ on what pattern it should form. Degeneration from created kinds gives no reason for the same lesions to be shared by exactly the groups that anatomy and retroviral insertions also group together.

Evolution is not only loss, either. Measured gains of function are covered in Genetic Information.

"Similar sequences reflect a common designer using common parts"

Common design accounts for shared working parts. Engineers do reuse components.

It does not account for shared broken parts, which is why this evidence is framed around damage. The reasoning is the one courts use for plagiarism: matching correct text proves little, while matching typographical errors in matching positions is hard to explain independently.

What the evidence shows#

Pseudogenes are disabled copies of genes that still work in related animals, so their former function is known rather than assumed.

The disabling mutations are shared according to relatedness: one set of lesions across all haplorhine primates for vitamin C, one set across all living birds for enamel, one across baleen whales, one across penguins. Independent losses — guinea pigs, individual bat lineages — look different, as independent losses should.

In at least one case the sequence was found by predicting where it should be: mammalian egg-yolk pseudogenes were located in the chromosomal positions that synteny with birds specified.

For the human-specific case, see Pseudogenes in the Human Genome. Related evidence appears in Endogenous Retroviruses and Vestigial Structures and Atavisms.