Endogenous retroviruses and the evidence for evolution

Source:  Endogenous retroviruses and the evidence for evolution    Tag:  common human viruses
Here is my description of the part of my research that has been misrepresented by the Reasons to Believe website. See this post for an explanation.

My paper concerns the regulation of a human gene by DNA derived from an endogenous retrovirus (ERV). An ERV is a viral sequence that has become part of the infected animal's genome. Upon entering a cell, a retrovirus copies its RNA genome into DNA, and inserts the DNA copy into one of the host cell's chromosomes. Different retroviruses target different species and types of host cells; the retrovirus only becomes endogenous if it inserts into a cell whose chromosomes will be inherited by the next generation, i.e. an ovum or sperm cell. The offspring of the infected individual will have a copy of the ERV in the same place in the same chromosome in every single one of their cells.

This happens more often than you might think; 8% of the modern human genome is derived from ERVs. Repeated sequences of this kind were formerly considered to be non-functional, or "junk" DNA. However, we're gradually finding more and more examples of viral sequences that appear to have some kind of function in human cells. For example, many ERV sequences play a role in human gene regulation. ERVs contain viral genes, and also sequences - known as promoters - that dictate when those genes should be switched on. When an ERV inserts into the host's chromosome, its promoter can start to interfere with the regulation of any nearby human genes. In the example that I researched, the ERV promoter has become responsible for most of the expression of a particular human gene in the large intestine.

Creationists and intelligent design advocates like to think that because some ERVs have useful functions in the human genome, they must have been deliberately put there by a creator / designer with that particular purpose in mind. Of course, no-one can explicitly prove that that is incorrect - it's not a falsifiable hypothesis, and therefore it's not science. What we can show is that ERVs provide evidence in support of the theory of evolution.

Let's imagine how ERVs would behave within a model of evolution by common descent. An ancient creature, let's call it the common ancestor of all modern mammals, is infected by a retrovirus that becomes endogenous. All of the animal's descendants (i.e. all mammals) would be expected to carry the same ERV insertion (ERV1) in the same chromosomal location.

Fast forward in evolutionary time. Different lineages have evolved and diverged from the original common ancestor and there are now many different types of mammal in existence, all carrying ERV1. A small rodent, let's call it the common ancestor of mice and rats, is again infected by a species-specific retrovirus that becomes endogenous. This is ERV2. In a parallel event in a different lineage, the common ancestor of all great apes acquires a third insertion, ERV3.

Moving forward again, a fourth ERV appears in some of these new-fangled human thingies that are running around in Africa, but not in their hairier relatives who will eventually evolve into modern chimpanzees. The early humans spread out, and a fifth and (don't worry) final ERV arises in a population that is isolated in a discrete geographical location. The infection does not spread to other human populations.

So what would we expect? Humans, chimps, mice and rats should all possess ERV1. The mouse and rat genomes will also contain ERV2, the virus that infected their common ancestor, but not the primate-specific ERV3, 4 or 5 insertions. All great apes will share an identical ERV3 insertion; all humans will also possess an ERV4 insertion that is not found in chimps or other apes. In addition, some, but not all, humans will carry an insertion of ERV5. The rodent-specific ERV2 insertion will not be found in any primate species.

Now that several genomes have been sequenced, we have begun to test these predictions. The patterns of ERV insertions observed in modern species exactly match the predictions made by the model described above. Some insertions are shared between humans and mice and represent truly ancient viral infections. Others are found only in primates, and not in other species, obviously derived from an infection of the ancestral primate species after its divergence from other lineages. More modern insertions are found only in humans, while the youngest ERVs of all are found in some humans, but not in all. We do not find any examples of ERV insertions shared by, say, humans and mice, but not by chimps. Insertions are always shared by all species, and only by those species, that have a common ancestor. ERV insertions therefore provide excellent support for the theory of evolution by common descent.

My particular favourite ERV is found in various primate species, and therefore must be at least 25 - 30 million years old. I compared the sequences and activities of the same ERV promoter in the human, chimp, gorilla, and baboon genomes. Despite some minor "single-letter" point mutations caused by DNA copying errors, the promoter had essentially the same function in all four species. I struggle to understand why any kind of designer would decide to use different codes to perform the same function in different species, but there it is. I hypothesised that the ERV was only allowed to persist (that is, its meddling in gene regulation didn't kill the first organism in which it inserted, which was therefore able to pass the insertion on to its offspring) because the incoming ERV promoter behaved in a very similar way to the original host cell's gene promoter. I wasn't able to do the experiments I wanted in order to investigate this point, but another group subsequently did, and their findings supported my hypothesis. That's what happens when you make and test falsifiable predictions.

I could go on and on about the role of ERVs in genome evolution, but this post is too long already. I would encourage anyone who's interested to search for "endogenous retrovirus" AND evolution in the PubMed database of peer-reviewed research papers. You won't find much in there about creationism or intelligent design. ERVs are also the subject of their own blog, a decent novel (although the sequel was disappointing) and a post on This Week in Evolution!