Comparing PreXMRV-2 gag sequence diversity in laboratory and wild mice

[anciendaze]

PreXMRV1 was shown to be present as an ERV in samples from the mouse with the xenograft. The source for preXMRV2 was never identified. The argument that it must have been present, because it came out in the form of XMRV was circular. This would be true whether it came from an ERV in a mouse or arrived as an active retrovirus in either mice or humans.

An active retrovirus is much more likely to recombine with similar ERVs. This changes the probabilities of origin considerably. Instead of a single mouse in a single experiment we now have to consider the probability of recombination in every mouse with both the ERV and an active infection.

This study strongly suggests there is such an active retrovirus in mice, and it is young, not ancient. Sequence homology points to a mixture of polytropic and xenotropic strains, as reported in other cases. We won't know the actual tropism until we have the complete virus to study. In particular we need to know the envelope to determine how it enters cells. There is good reason to believe it can use XPR1 receptors, which humans have, but most mice lack.

The story is not over.

 

[Bob]

I've just had a look at the Paprotka study again, to refresh my memory.
Here's are some bits of info about the PreXMRV viruses that I wanted to highlight.

First of all, it says that PreXMRV-1 is replication defective.
Is anyone able explain to me how a virus can infect, if it can't replicate?

I've not followed the discussions re Paprokta closely. The following two paragraphs are worded strangely. They don't say that they detected PreXMRV-1 in the mouse strains, but only that the complete sequence of PreXMRV-1 was determined from various sources, including mouse strains. Could anyone explain the significance of this?

"However, we did find that early xenografts contained a previously undescribed XMRV-related provirus that we have named PreXMRV-1 (Fig. 2B). The complete sequence of PreXMRV-1 was determined from the early-passage xenografts, the NU/NU and Hsd strains, and the CWR-R1 cell line (GenBank accession no. FR871849)."

"Overall, these results indicate that PreXMRV-1 is an endogenous murine provirus that is present in the NU/NU and Hsd strains"

Other sentences in the paper do say that the mouse strains 'contained' PreXMRV-1, but after the PACE Trial, I'm very wary of the use of wording in published papers, and I'm confused by the lack of consistency. For example:

"...only the Hsd and the NU/NU outbred nude strains contained PreXMRV-1..."


With regards to PreXMRV-2, the paper says the following:

"Six of the 15 mouse strains contained PreXMRV-2, but only the NU/NU and Hsd mice contained both PreXMRV-1 and PreXMRV-2"

So the paper says that both PreXMRV-1 and PreXMRV-2 were found in mice strains.
I seem to remember noticing some discussions where people said that this isn't the case, but I didn't read the details.
Is anyone able to explain if I'm missing some info here please?

 

[August59]

Good reading anciendaze and Bob - I've always felt like something is there, but whatever it is is very evasive and make be the one that sort of breaks the rules as far as virus discovery or detection goes. Although I do believe a lot of what was found to be positives by Dr. Mikovits, Dr. Alter and D. Lou and others was contamination, I do not believe all of it was! I believe the variations in techniques, primers, assays and other possible (and probably very minor) variables led to the differences in what they reported in their respective papers. But, I'm sticking to my gut feeling that there is something there that is unique whether it as new virus or a genetic mutation that changes the normal way a virus is processed in our bodies. A case in point is the people with a certain gene or receptor that makes immune to being infected withe HIV virus.

It is my hope that the Lipkin study, while may not find XMRV, will present enough evidence that their may be a new virus, a muted virus or a virus is present in ways that it should not be by scientific standards due to what almost has to be a genetic defect or mutation. Any of these would renew extreme interest within the research community and therefore push the visiblity of ME\CFS even higher. Anything that does this will have an exponential impact on treatments for us down the road. I really do hope we get to see the Lipkin paper around July 30th as someone recently posted.

 

[anciendaze]

You have sharp eyes Bob.

No, a replication-defective retrovirus ordinarily cannot infect. It may, however, be transcribed into (defective) RNA and proteins. These are what I'm guessing the researchers were assuming managed to recombine. As for what triggered the transcription of defective sequences, anyone can make a guess. Part of the lore of ERVs is that they have a mysterious ability to be resurrected from defective sequences. If you look at all the possible causes in the literature you could get the impression these are exactly the same as causes which activate a fully replication-competent provirus which just happened to lie below the threshold for detection. That is what I believe.

The preXMRV1 sequence was only determined in part from actual sequence data. Since nucleic acid sequences for some parts of the genome have not been entered in Genbank, I'm assuming they extrapolated backwards from sequences found in XMRV. (This came to my attention because I was interested in the envelope, which contains an immunosuppressive domain that is functional in transgenic mice with approximately human immune systems, as shown by other research. This piece alone could cause health problems. Even if it arrived through pseudotyping from the envelope of another virus it should be tracked down.) Nobody seemed particularly bothered by the circularity of this reasoning.

With respect to preXMRV2 the language is more definite than warranted. What they are saying is that it might be in mice from the strain of nude mice, not that it was detected in the particular mouse where recombination was postulated. If an ERV arrives by vertical transmission through the germ line it will be present in every normal nucleated cell. and should be detectable. An active retrovirus could also be transmitted from mother to offspring while nursing, but not be in the germ line.

At this point we have to deal with a peculiarity of nude mice. They have defective immune systems, so it is not a good idea to expose them to other laboratory animals which might carry active infections. They are also defective w.r.t. nursing offspring. Because neonates generally benefit from milk from a mother with a healthy immune system nude mice are typically nursed by brown mice. It is hard to keep them alive otherwise. PreXMRV2 has definitely been found in brown mice, where it was assumed to be endogenous. If anyone had checked the pattern of inheritance they might have detected indirect evidence of an active retrovirus. Vertical transmission via milk can be confused with transmission through the germ line.

Other retroviruses have been demonstrated to transmit via nursing, but detection of virions in milk is difficult. Animals with active infections typically have antibodies in milk which are likely to inactivate free virions. You can also find maternal cells, including stem cells, in milk. In the case of MMTV we know the next link in the chain takes place at dendritic cells in the lining of the neonate gut. These then migrate to lymph nodes where they present antigens to activate other immune cells. If infected maternal immune cells pass antigens, including complete virions, directly to neonate dendritic cells the mystery would be solved -- transmission of the virus might not require free virions at all.

There are three oversights in the reasoning which led to this claim: 1) the presence of preXMRV2 fragments was assumed to mean this was an ERV, because nobody noticed an active infection; 2) nobody was very worried about active infections which left animals asymptomatic; 3) nobody seems to have worried about thresholds of detection, and the possibility infections might lie below this.

If you sample past literature on the subject of endogenous sequences, as I have, you can easily reach the conclusion that, had the forces in the debate been arrayed a little differently, a sequence now solidly listed as endogenous might have ended up in a different category. Both MMTV and JSRV are known to be present as both exogenous viruses and endogenous sequences. Make detection of active infections a little harder and they would be called ancient ERVs, assuming you did not compare the distribution of these sequences in a population carefully.

 

[jace]

In the comments section of O'Keefe's blog, someone asked a question:

Could it be that other mice are infected but again the ability to detect them is beyond science at this time?

When you say a young clade how old would that be?

and O'Keefe answered

sorry not to get back to you earlier; I have been working all weekend - in answer to your first question, I would have to say yes, because how can we know for sure they aren't there if we can't detect them. It really is a matter of probability however - if the correct tissues and controls are used, then chances are if it were there, it would be detected. In answer to the second question - without going back and doing some reading (sorry no time right now!)I honestly don't know. We are talking in evolutionary terms, meaning the phylogenetic "tree" analysis doesn't show a lot of evolution from the original sequences. This would be comparing the patient (ie from Lo et al.) and cell line sequences as well, that they mention in the paper. When I have done this kind of analysis on human genes, recent could be considered a few million years (its all relative). In this case, to be more specific, we might have expected that samples from different mouse strains and cell lines and humans would be more similar to other sequences from the same source - ie the preXMRV-2 sequences from humans would be more similar to each other than they are to a wild-mouse caught in Germany. And like-wise, sequences from mice of a particular strain would be more similar to each other and separate out on a phylogenetic tree. In addition, not all wild-mice have the virus, again suggesting if it were truly old, then more of them would be positive. Further sequence analysis (ie longer reads) may help to differentiate the viruses further. To caveat these comments however, I am no specialist in evolutionary virology - these are just my thoughts.

http://okeefe-lab.blogspot.co.uk/2012/07/next-gen-sequencing-reveals-that-pre.html#comment-form

 

[anciendaze]

Just in case anybody missed the point of my previous post, I'll make something implicit above explicit here. You can argue that all this is merely evidence of activity by mouse viruses of no relevance to human health. I've actually supplied a speculative outline of how this might work. The reason you don't hear a great deal about such a line of reasoning at present is that it requires admission that virologists working with common mice have misinterpreted evidence of active infection in their most common laboratory model of retroviral infection in mammals over a period of many years.

I'll throw another speculative fox in the hen house while I'm here. In evolutionary terms the loss of the XPR1 receptor is quite recent, and occurs in virtually all mice commonly associated with human civilization. Ignoring mice, XPR1 receptors count as highly conserved in many species. This anomaly deserves explanation. One possibility is that carrying a pathogen which is more dangerous to predators found in association with human civilization than to the mice themselves is advantageous to mice as species (though not to individual mice killed by predators.) In this case, the spread of these viruses might not have begun until people domesticated other animals and began living in fixed locations. In this scenario humans could be bystanders caught in an evolutionary arms race between predators and prey.

 

[natasa778]

Do we know anything else about XPR1 apart from it being used by (these) viruses? What use is it, or was, for the host? Any known ligands?

 

[natasa778]

just answering my own question, as none are known yet but still it seems XPR1 is very important to cell function

from http://labs.fhcrc.org/miller/research.html

In later studies we focused on how XMRV might be involved in CFS, and found that XMRV could kill neuroblastoma cells, indicating a neurotoxic activity of XMRV that might be involved in CFS. This toxic effect was mediated by interaction of the virus with its cell-entry receptor, Xpr1, which we recently found was an atypical G-protein-coupled receptor that regulated cAMP levels in the neuroblastoma cells. XMRV infection blocked Xpr1-mediated signaling resulting in reduced cAMP levels and cell death.

... However, we continue to study the role of Xpr1 in cell biology, and are looking for possible natural ligands that might regulate Xpr1 activity. Xpr1 orthologs are present in many organisms, including plants, animals and fungi, and XMRV has provided a window into what may be an important regulatory pathway in a wide range of living things.

... and still mice sacrificed it!
btw DM might want to rethink the bit on the origin of XMRV in the light of new data. Just saying ...

 

[Bob]

I'm still trying to get my head around the Paprotka paper, and this is what I've understood so far...
But bare in mind that I might be misunderstanding it...

I've been studying Fig. 1 and Fig. 2 in Paprotka, and it seems that PreXMRV-1 was found (as genomic DNA) as a complete genome (minus some of the LTR info), in some of the Xenograph passages, assuming that they've carried out their methodology and analaysis correctly.

PreXMRV-1 was detected as genomic DNA in the NU/NU and Hsd strains, but no active virus was detected, so it is assumed to be endogenous.

However, if I'm understanding it correctly, it seems that only a tiny section of PreXMRV-2 was detected in any Xenographs. The PreXMRV-2 sequences that they actually detected only included one LTR sequence and a gap (the 24-bp deletion in the gag leader characteristic of XMRV), and not the gag, pol or env sequences. So PreXMRV-2 wasn't really detected in the study at all (only a tiny portion of it - one of the LTR sequences). So they are making an assumption, and a deduction, that PreXMRV-2 was involved in creating XMRV. Almost half of PreXMRV-2's genome is included in XMRV, but very little of PreXMRV-2 was detected in any of the Xenografts (just the LTR info), and PreXMRV-2 isn't present in 22RV1.

It's interesting to note that PreXMRV-2 is encoded as mouse genomic DNA in the DBA/2J genome, so I think that means that it must be an endogenous virus.

PreXMRV-2 is an endogenous virus because it is coded for in the DBA/2J genome. I don't think that any active PreXMRV-2 viruses were detected in mice or xenografts - just genomic DNA.

The paper says:
"In a search for proviruses that might contain XMRVspecific sequence features, we found a second previously undescribed endogenous provirus that we named PreXMRV-2 (Fig. 2C)."

However, I can't work out where they got their genetic info from, because I can't see anywhere in the paper where they have sequenced PreXMRV-2, except for the existing recorded mouse genome. I think they were making an assumption that PreXMRV-2 is a virus purely based on the overlap of genetic data between mouse genome, XMRV and both the preXMRV viruses. If this is the case, then they deduced that PreXMRV-2 exists, but did not detect active viruses. So they looked at a strip of mouse genome (DBA/2J), and saw that half of it was incorporated into XMRV, and they had detected a tiny portion of it in the Xenographs, and then deduced that the rest of the strip of genome was part of a single endogenous virus.

I suppose that a similar process was carried out for PreXMRV-1, if no active virus (viral RNA) was detected, but at least they found most of the PreXMRV-1 genome in the xenografts, as genomic DNA.

Phew! Now my head hurts! I'll come back and have a look at this later, and see if I've made any glaring mistakes. And I'll have another look at the paper to see if I've missed anything vital.


Interesting additional info:
This is my interpretation, and it could be very wrong, but there also seems to be another preXMRV endogenous virus (it's an XMRV-like sequence of genomic DNA) (see Fig. 2A), in xenograft passage CWR22R 2152 (or, to be precise, I think it's a lapsed androgen-independent tumor), which could potentially be significant, and I don't think that they've discussed this in the paper... It is missing about four-fifths of the XMRV gag gene, and is missing about a fifth of XMRV pol gene... But I don't think they've discussed any potential significance of this, but have ignored it... I'll have another look at it later.

 

[Bob]

If anyone wants to read deeper, I've found this paper:

Characterization, Mapping, and Distribution of the Two XMRV Parental Proviruses​

Oya Cingöz, Tobias Paprotka, Krista A. Delviks-Frankenberry, Sheryl Wildt, Wei-Shau Hu, Vinay K. Pathak, and John M. Coffin
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3255884/

Have fun!!!

 

[anciendaze]

You need to be careful, Bob, to distinguish searches done entirely in silico from laboratory work directly pertaining to XMRV in nude mice. I can tell you from experience that active databases accumulate bad data. I've seen some attributions change during this controversy even though the associated sequences remained the same. Once preXMRV2 was labeled endogenous in one place that unchecked assumption propagated all over.

 

[Bob]

You need to be careful, Bob, to distinguish searches done entirely in silico from laboratory work directly pertaining to XMRV in nude mice. I can tell you from experience that active databases accumulate bad data. I've seen some attributions change during this controversy even though the associated sequences remained the same. Once preXMRV2 was labeled endogenous in one place that unchecked assumption propagated all over.

I don't completely understand what you mean, anciendaze.

The paper says that they are only assuming that PreXMRV-1 is an endogenous retrovirus, based on the finding of an almost complete copy of gDNA in the xenografts, and also by making deductions. There doesn't seem to be any existing mouse genomic data to confirm that it is an endogenous retrovirus.

But PreXMRV-2 seems to be an established section of mouse genome (DBA/2J). I don't know its history, but there seems to be plenty of research based on DBA/2J. They found this DBA/2J genomic sequence in the two mice strains associated with the xenografts, but they haven't detected PreXMRV-2 (or the associated DBA/2J sequence) in any of the xenografts, so they are making a lot of assumptions about PreXMRV-2 being involved in the creation of XMRV.

My understanding is that they are only able to assume (and to partially deduce, based on the partial evidence) that both, or either of, PreXMRV-1 and PreXMRV-2 are involved.

 

[Bob]

OK, summing up my earlier post, my current understanding of the Paprotka paper is as follows...

Almost the complete PreXMRV-1 genomic DNA (gDNA) was found in some of the xenograft passages.

PreXMRV-2 wasn't found in the xenografts (or only a tiny fragment was found, which wasn't enough to determine that it was PreXMRV-2).

They base their model of PreXMRV-2 on an identical (or near enough identical) existing mouse genome data, so they have made deductions, based on assumptions, that this mouse DNA sequence is an endogenous virus sequence, and that it contributed to making XMRV.

Only a very small section of PreXMRV-2 was found in the xenografts (only an LTR sequence, and the gap: the 24-bp deletion in the gag leader characteristic of XMRV.)

PreXMRV-2 (gDNA) was detected in a variety of mouse strains, but it looks like they may have only been testing lab mice (I'm not certain about this, as I haven't looked closely at this detail yet.)

Also interesting is what looks like another preXMRV sequence, which they haven't commented on, in Fig. 2A.

 

[Mula]

Paprotka assumed that the two prexmrv viruses were endogenous but Mayer found that prexmrv2 has diversity which an endogenous retrovirus would not have. Prexmrv1 in the initial paper was not found intact from a single source but was complied from three sources. A variety of unknown viruses will be available which could establish the other half of xmrv.

 

[Bob]

Paprotka assumed that the two prexmrv viruses were endogenous but Mayer found that prexmrv2 has diversity which an endogenous retrovirus would not have.

Yes, it demonstrates that Paprotka was based on limited knowledge.
They do admit that they make some assumptions, and it should never have been regarded as a conclusive study.

Prexmrv1 in the initial paper was not found intact from a single source but was complied from three sources.

Oh, that's a very interesting detail, that I don't think is explained in the paper.
Do you happen to know where I can find the details for that, please Mula?

A variety of unknown viruses will be available which could establish the other half of xmrv.

I think there are a myriad of potential explanations for how XMRV was created.
But XMRV does have an affinity to human tissue, so it does seem plausible that it was possibly created in this cell line. (Can XMRV infect mice as well as human tissue? I've forgotten! :thumbdown

Otherwise it could have jumped from lab workers to the cell line...

Or another possible explanation, based on my limited understanding, is:

As far as I can see, the authors haven't explained the XMRV-like sequence of genomic DNA, in xenograft passage CWR22R 2152 (see Fig. 2A). They seem to have ignored this detail, but it looks to me like it could be possible that this XMRV-like partial viral genome could have been a mouse or human retrovirus that entered the cell line at a late stage, and could then have recombined with another one of the preXMRV viruses, to form XMRV.

 

[anciendaze]

We are getting into a level of detail which involves considerable speculation about actual findings and sources. I will try to answer without claiming too much specific knowledge.

(Can XMRV infect mice as well as human tissue? I've forgotten! :thumbdown

It can definitely infect mice retaining the XPR1 receptor, that has been established. It also shows evidence of using some other route of transmission which has not been pinned down. It is difficult to get it to infect most mice, not impossible. The term xenotropic for this group of viruses was probably a misnomer to begin with as they have a wide host range.

In general terms, showing a sequence is definitely endogenous means showing it came through the germ line. A proviral sequence from an active infection will be limited to infected cells of particular types. An ERV will be present, though not active, in every nucleated somatic cell. Furthermore, it will be integrated into chromosomes at the same location. At least one parent should also have the ERV.

Showing this beyond doubt is a lot of work. Most researchers will not do this if the sequence closely resembles other ERVs. In fact, finding the sequence in both maternal and offspring DNA is generally enough to convince them they are seeing an ERV. Most of the time this is correct because only a small part of the genome changes over short times.

A major exception occurs when there is an active retroviral infection. The problem here is that most researchers thought they knew there was no such infection. This allowed them to mistake vertical transmission of an infection via milk for vertical transmission of a gene via the germ line.

Cleaning up after this long-standing misconception will be a problem. While most genetic studies probably did not depend on genes inserted by recently active retroviruses the fact that there are exceptions should undermine confidence in them. Other studies, on HIV, have assumed there was only one active retrovirus involved in a study. If there were more than one, recombination would have a major effect on diversity of quasispecies and drug resistance.

 

[lansbergen]

A major exception occurs when there is an active retroviral infection. The problem here is that most researchers thought they knew there was no such infection. This allowed them to mistake vertical transmission of an infection via milk for vertical transmission of a gene via the germ line.

Cleaning up after this long-standing misconception will be a problem.

Milk is not the only option. There can be other transmissionroutes during pregnancy and birth.

Furthermore newborn animals can make the mistake to suck the vulva instead of a nipple. I have seen that happen.

 

[natasa778]

Milk is not the only option. There can be other transmissionroutes during pregnancy and birth.

Absolutely, have a look at this one (btw the nearly 100% transmission rate relates to second pregnancies)

http://online.liebertpub.com/doi/abs/10.1089/0882824041310586

Overall, 99% vertical transmission occurred in pups of infected mothers. Twelve percent of mid-gestational and 39% full-term fetuses were PCR positive. We have established that, if mothers are infected with ts1 virus at 72 h after birth, then nearly 100% vertical transmission occurs, via in utero, intrapartum, or breast milk.
Thirty-nine percent transmission occurred in utero alone.

 

[anciendaze]

Milk is not the only option. There can be other transmissionroutes during pregnancy and birth....

I never said there were not. It only takes one such example to illustrate the error in reasoning.

I used milk as my example because of the work that has been done on MuLV, MuMTV, FeLV, etc. You can also find work on transmission of HIV by this route. The problem there is that you can still find scientists denying that HIV transmission takes place this way because of antibodies in milk. They are ignoring possible transmission by infected maternal cells in the mixture, which are really hard to find, and concentrating on free virions. I believe this is an important gap in thinking which has helped to make it possible to go so long without noticing active infections.

Having reached this point in the discussion I want to raise an issue which has turned up when HIV+ patients are pitted against ME/CFS patients in funding battles. We have a parallel animal model, though not an exact parallel, in the case of FeIV and FeLV. Cats may be infected by either retrovirus alone. The infection can still be lethal, though the cat may live quite a long time in a state which allows them to spread infection. Cats infected with both retroviruses decline more rapidly, and mortality rates are much higher. There are similar variations in human response to HIV infection, largely unexplained.

Is it coincidence that the "drug cocktails" used to arrest HIV infection are likely to contain one or more drugs which also affect XMRV infection?

(Keep in mind that the current definition of XMRV refers to the recombinant from a laboratory. We have yet to study the complete preXMRV2.)

 

[barbc56]

Ancientdaze, I just want to clear up something you said in your post. You may have meant this but people can not catch feline leukemia from cats.

Barb C.