What is the normal course of events leading to endogenous retroviral sequences in most mammals? It starts with infection by an exogenous retrovirus, naturally. While endogenous retroviral sequences may be recombined, there is no reason to believe they appear by spontaneous generation.
After a sequence has been incorporated into germ-line cells it may still fail to be fully endogenous. Activation of the inserted genes may have a deleterious effect on differential reproductive success of the host. This may even prevent offspring from being born. To become widely distributed in the host species their evolutionary impact must at a minimum be neutral. Some might even be beneficial in terms of reproductive success, even if they result in the death of hosts after reproduction. Oncogenes related to breast cancer might fit this description, but at this point we don't understand the etiology of breast cancer in humans.
If the inserted genes are not actively expressed, and don't cause some malfunction in gene expression on the same chromosome, all mutations in them are evolutionarily neutral. It doesn't matter whether they encode some enzyme needed to transcribe RNA into DNA, or the secret passphrase on the home wireless network of the secretary of HHS, if this encoding is not expressed, it is nothing but evolutionary junk. In time the random walk of mutation without selection will result in sequences which cannot be activated. The probability of a sequence which has wandered away from an active form returning to that form on its own is very, very small. After millions of years all such viruses become extinct. The two-stroke engine of variation and selection drives evolution, a one-stroke version does not.
Is there any way for an ERV to escape this fate? Yes, and we know one such example in considerable detail. Murine Mammary Tumor Virus (MMTV), a beta retrovirus, is both an endogenous retrovirus (ERV) and an exogenous retrovirus in mice. Even if an endogenous sequence is defective it can be activated in the same way as an exogenous sequence, the LTRs tend to be the last functional parts to degenerate. Once activated the virus is again subject to selection pressures. It also mutates and recombines at much higher rates than it does while passively ensconced in a chromosome. (Consider the difference between estimated rates of mutation in human endogenous retroviruses (HERVs) and the perennial battle between public health and new strains of RNA virus that cause flu.) But, how does this activation take place?
The easy way to activate a specific ERV, and nothing else, is infection with an exogenous retrovirus which is very similar. It need not be identical. Once this process is set in motion the reactivated ERV can once again mutate to converge on the optimum sequence for that host. If the HERV is defective, the exogenous virus may serve as a "helper" providing necessary functions. Since the ERV sequence is present in virtually every cell of the host this offers a perfect camouflage for hiding from host immune response. (This is called molecular mimicry, but I'm not really sure the name is appropriate. A virus may even incorporate host sequences directly. This is less like mimicry than the defensive camouflage of a hermit crab.)
In the case of MMTV we see that the interaction between endogenous and exogenous sequences is so active that new sequences are being inserted as fast as old sequences degrade into extinction. The infection has reached equilibrium, and may remain both endogenous and active for an indefinite time. This process appears to have happened many times, in many species.
When I hear that humans are biologically unique in having no active beta or gamma retroviruses I become uncomfortable. I've lived through a period when humans were unique in having no active retroviruses whatsoever. We all know how that turned out.
Part of the problem is that these viruses are not merely present in mice, where we have a disproportionate weight of data. They are present in a variety of mammals, including domesticated animals. They also appear to be present in most primates, though our information about some is spotty. Some, like Gibbon Ape Leukemia Virus (GALV), are so similar to murine leukemia viruses that they can be confused, and have been misidentified in the past.
If we had better information about ERVs in domesticated animals I believe we would see that there are commonalities which can only be explained by active viruses, even in cases where we have failed to detect them. Remember that all domesticated animals except dogs have been domesticated in roughly the last 5,000 years. This is an evolutionary experiment on a grand scale. It is also a massive change in human environments.
If humans are surrounded by such viruses, even in animals consumed as food, and related species are definitely infected, there would have to be some strong reason to believe humans are immune to infection. To my mind no such reason has been presented.
What we have seen are a long list of findings indicating that HERVs may become active in MS, SLE, RA, etc. There are also many findings of beta retroviral sequences related to breast cancer. Without an exogenous virus to start the process this becomes a distinctive anomaly.
What is true beyond argument is that it becomes increasingly hard to distinguish exogenous retroviruses, or those from associated species, from endogenous sequences. We have plenty of evidence that the ancestors of humans were repeatedly infected by viruses with closely related sequences. These were eventually defeated by host immune systems, but only after they had infected a large part of the population long enough to enter the germ line. At a later date another retrovirus which was similar, but not identical, would start the process anew.
The result is that humans now possess dozens of sequences from beta and gamma retroviruses which immune systems must tolerate. If this fails the result is autoimmune disease. What triggers such disease? Common answers are it could be many things. This translates into an indirect admission that "we don't have a clue how this happens."
If you find virions resembling the progenitor of a family of HERVs in those with these diseases it is likely an active infection by an exogenous virus started the process. Without active infection to begin with, the process of mutation under selective pressure which reconstructs an optimal sequence is unlikely to take place even once in human history because there is no selective pressure on passive sequences.
At this point those I dub sequence technicians rush in to say they looked for that identical sequence as an exogenous virus and failed to find it. Do active infections by viruses which carry genetic information in RNA mutate rapidly? Are they aware of another retroviral quasispecies, currently causing considerable concern, which has strains only 70% homologous to each other? Some tools are simply inadequate for this search.
Concerning those HERVs homologous to beta and gamma retroviruses I may be told human immune systems have become sufficiently competent to stop new infections of those types. Where others see a solid wall of defeated enemies I see a thicket of endogenous sequences which must be tolerated. This, in turn, causes me to remember Br'er Rabbit saying "Please, don't throw me in that briar patch!" Where else would a retrovirus hide?
Is anyone going to tell me "It wouldn't be fair for a retrovirus to do that"? Could this possibly mean "It wouldn't be convenient for us if a retrovirus did that"? Can you point to any natural law requiring human convenience?
Blog entry posted by anciendaze, Nov 15, 2011.