Theyre figuring out what the heck is going on in the saga of ERVs and ALS!
1-- Of the many ERVs in your genome, the ones they found activated in the brains of people with ALS are HERV-Ks, at two positions: one on chromosome 7 and one on chromosome 10. ERVs degrade a lot over time, but they could find transcripts for all the major retroviral proteins, gag, pol, and env.
Why is this important?
At first, they could only find activity from the chromosome 7 locus, only in some patients. HERV-K activity from chromosome 10 is a new one, for me. There might not be only one HERV-K that can wake up and cause ALS.
Also, they only looked for pol before. It seems the previous work was focused on the enzymes of ERVs, including reverse transcriptase, possibly because RT activity has been detected in ALS patients. Theyve even tried giving people with ALS RT-inhibitors. But in this study, they looked for mRNA from all three major retroviral proteins.
2-- The Env protein appears to be the one causing trouble. Again, before scientists were focused on RT. In this study, they looked for the Env protein, and found it in ALS brains. Furthermore, whether they put active HERV-K genomes or only the Env gene into neuronal cells, both were cytotoxic to the cells.
In the absence of any other environmental/physiological/anything, these HERV-K Env proteins are enough, by themselves, to cause neuronal damage.
3-- This knowledge could be translated into an Env-expression based small animal model for ALS. They made a transgenic mouse that would express HERV-K Env in neurons. The prefrontal cortex of these animals was fine. It was the upper and lower motor neurons that died. Mice progressively lost motor function.
ALS. "Your mind is fine, while your body is collapsing around you".
4-- What about TAR DNA binding protein 43? In the previous study, TDP-43 was a confusing issue. It was playing a role in this ERV-ALS story, but they couldnt quite figure out where it fit. But now they know HERV-K Env expression alone is enough to cause neuronal damage/ALS-like disease in mice. Whats up with TDP-43?
Point #2? How they could put HERV-K Env in human neurons and see damage? They tried the same experiment, but but TDP-43 in human neurons. Thats it. Nothing else. Guess what happened?
They saw increased expression of HERV-K Env, and cytotoxicity.
They also found *exactly* where TDP-43 was probably binding in the HERV-K DNA to make these unwanted transcripts-->proteins.
This paper has answered a lot of questions about the relationship between ERVs and ALS (some ALS, not enough patients in this paper to definitively say 'ALL ALS!'), but there is still a lot to do before this translates into any therapies/preventative measures for patients. The main questions for me involve TDP-43, not the ERVs. 1) Why does TDP-43 start snuggling up to ERVs? Whyyyyy? Why would it start doing this?? 2) How the hell can you prevent/stop TDP-43 activating ERVs? Not only do scientists have to figure out how to do it, they then have to figure out how to get this therapy into the CNS. It is *hard* to get any kind of drug in there-- its going to suck.
So, good news/bad news with ERVs and ALS-- Good news is these folks are really starting to figure out what is going on with this disease. Bad news is, its not RT causing trouble, which explains why RT-inhibitors did nothing for ALS patients. RT-inhibitors would have been a nice, already tested/approved drug. TDP-43-Env relationship will need an entirely new approach.
And, repeat after me, folks:
"The vast majority of ERVs are junk. JUNK DNA. This is a good thing. When junk DNA is accidentally reactivated, BAD THINGS HAPPEN. Not magic unicorn rainbow things. BAD THINGS."
A question for anyone who knows the answer, from a non-biologist. I am trying to understand, of the 100,000 or 200,000 (I have seen both numbers) fragments of ERVs, or of the ~230 (as of 2005, refc below) full-length ERVs, how many are (1) the result of separate infection events, versus (2) just being duplicates of the first class? Does each infection and its duplicates get classified as a whole new "family"?
Here is the meatiest reference I have found so far. It is ten years old :" Recent evaluation of the human genome sequencing data revealed that about 9% of the human genome is comprised of elements with long terminal repeats (LTRs) (LTR retrotransposons) (36, 43, 84) comprising over 200 families (30). The majority of these LTR elements, however, lack sequence similarity to retroviral genes within their internal region or constitute solitary LTRs. About 40 families identified so far have at least some members that show discernible homology to coding regions of retroviruses, but most of them have not yet been analyzed in depth (47, 74). These families are grouped into three classes based on the sequence homology of their pol regions with the pol genes of exogenous gammaretroviruses, betaretroviruses, and spumaviruses. They comprise around 200,000 entities (36), including about 230 full-length proviruses. " - -
Seifarth. et al., J. Virol. 79 ,341-352 (2005)
Thanks in advance...
Almost certain nobody knows the answer to your first question, it would take a rather tremendous amount of work to get a good estimate, especially for more ancient ERVs; the older they are, the more fragmented they get, and the more difficult to piece together what the original provirus looked like.
For HERV-K HML-2, this analysis has been done, as there aren't that many loci, and they are fairly intact -there are ~100 full length/partial proviruses (10 times that many solo LTRs) and 12 of them are the result of duplications, from maybe 4 unique integrations. See figure 2 of this paper: http://www.retrovirology.com/content/8/1/90
In terms of HERV 'families,' that's a somewhat contentious area, but the answer to your specific question is no. HERV 'families' are not well defined (they certainly do not correspond to viral families, as all retroviruses, endogenous or not, belong to the family Retroviridae), but generally correspond to large-ish, distinct clades of HERVs, most of the time with more than one unique integration. For example, the HERV-K HML-2 clade is sometimes considered a 'family,' sometimes just a sub-group of the larger HERV-K 'family,' which contains, at a minimum, all 10/11 HML clades. Either way, there are shed loads of independent integration events within the group, spread over the course of millions of years of primate evolution.
I should note that, although your question about duplication/vs integration is hard to answer exactly, I can say that there are a *lot* of unique integrations, it's certainly not completely dominated by duplication events. Once a HERV group has completely lost replication competence, you would generally expect the duplication/unique ratio to increase over time, as at that point the only way to increase copy number is duplication.
Also, there's the related question of how many ERV integrations are formed from infection of exogenous viruses vs intracellular retrotransposition, i.e., transcription->reverse transcription->integration, skipping virion budding and cellular entry. That problem has been explored in a bit more detail than the duplication question, see here: http://mbe.oxfordjournals.org/content/22/4/814.full and here: http://www.pnas.org/content/109/19/7385.full
Thank you for your prompt and comprehensive replies. I read through the references you provided. Your summaries were very helpful in stating the relevant conclusions. Also, I didn't know the right words to put it in, but what I was actually wondering about was infection of exogenous viruses vs intracellular retrotransposition,, which you addressed in your second reply. Thanks again.