ERVs and Multiple Sclerosis, #4

I wasnt super thrilled about the last ERV-->MS paper I wrote about, but I really like this new one:

Human endogenous retrovirus type W envelope expression in blood and brain cells provides new insights into multiple sclerosis disease.

Those of you who have been following the XMRV fiasco will find this intro paragraph very interesting:

When the first descriptions of retrovirus-like particles with reverse-transcriptase (RT) activity in leptomeningeal and macrophage cell cultures from patients with multiple sclerosis (MS) were published they were thought to be related to a new human T-lymphotropic virus (HTLV) supposed to explain analogies between HTLV type 1 associated myelopathy and MS. After preliminary biological characterization of this MS-associated retroviral element (MRSV), molecular analysis of its genome revealed quite a complex picture as the MSRV genome had endogenous counterparts in human DNA featuring a Human Endogenous Retrovirus (HERV). Instead of the expected HTLV virus, it represented an element from a previously unknown HERV family, now named HERV-W.

Transport yourself to 1989. You take brain and blood cells out of MS patients... and find reverse transcriptase activity. You can find RT activity and viral genome associated with viral particles. HTLV and HIV are all the rage, so youre thinking you found a new retrovirus. Other labs in other parts of the world can replicate your findings. Other labs around the world figure out a potential mechanism for disease-- the envelope protein is pro-inflammatory. Its probably the viral protein causing trouble. Theres only one problem: When you do all the proper controls for your PCR experiments, you are finding sequences that look like your 'new' virus in human DNA. There is no human genome project yet, but we know that there are endogenous retroviruses in human DNA. So you dont conclude that there is a massive, world-wide conspiracy against you, you conclude that the offending agent isnt a 'new' retrovirus at all, but a million year-old endogenous retrovirus that has somehow become activated in MS patients, and might/might not be directly contributing to disease. So you spend the next couple of decades trying to figure it out.

I think this paper, while not a sexah Nature or Science or Lancet or whatever paper, is an honorable contribution to the foundation of work the 'MSRV' researchers set in 1989. Its practical and thorough.

MS patients + appropriate controls (not only healthy controls, but control samples from patients with other neurological and non-neurological diseases):

Samples comprised (i) other neurological diseases (OND): 20 patients with epilepsy, chronic polymyositis, primary cerebral tumour, sciatica, Guillain-Barré Syndrome, stroke, multisystem atrophy, facial palsy, amyotrophic lateral sclerosis (ALS), cerebral metastasis (lung cancer), Leigh's disease, traumatic medullar lesion, cerebral abcess (listeria), together with eight patients with chronic inflammatory demyelinating polyneuropathy (CIDP); (ii) non-neurological diseases (NND): 15 patients with chronic hepatitis B virus infection, 15 with chronic hepatitis C virus infection, 30 with systemic lupus erythematosus (SLE); and (iii) healthy blood donors: 60 HC.

They also used multiple avenues of blinded detection: ELISA, quantitative real-time PCR, RT-PCR

What they found: In the pilot study, 23 of 29 MS patients were positive by ELISA (had HERV-W Env expressed in their blood cells). Other groups from other locations got 59 of 74. Some of their diseased controls, like the CIDP patients, for instance, were also positive (5 of 8). The healthy patients were used to set baseline, but none of the diseased (hepatitis B patients, for instance) were Env postitive.

This is where the diseased controls are important-- The lupus, hepatitis, etc controls were NOT HERV-W positive. Because, an alternative conclusion could be that any chronic disease leads to some kind of epigenetic disregulation and Env is expressed, rather than Env being expressed is causing disease. If the patients with other chronic diseases do NOT have Env present, then there is more weight towards the hypothesis that HERV-W is causing trouble in MS patients, not the other way around.

So in summary, 80% of the MS patients had HERV-W Env protein.

In quantitative real-time PCR for both RNA expression and DNA copy number, 40% of MS patients were increased relative to healthy controls, whereas 4% of the blood donors were increased.

While this is all find and dandy, its a meaningless trick without some connection to the brain.

As it is relatively difficult to get brain samples from living donors, the investigators looked at the brains of 8 MS patients who had donated their bodies to science after death. They found that HERV-W was expressed on the macrophages and the activated microglial cells present at active MS lesions in the brain. They got the same results with three different anti-env antibodies, and different macrophage markers, in all eight brains.

This paper made a direct, clear connection between what we had been observing in blood samples with what is going on in the brain. Its the macrophages and microglial cells (macrophages that are specific to the brain), that are expressing HERV-W in the brain. We can test for the 'brain' phenotype by looking in the blood. You dont have to go into CSF. You dont have to wait and look at brains when the patient is dead. They also came up with this theoretical model of disease.

And they mentioned something I have not heard before... Anti-HERV-W antibodies are in clinical trials for MS patients...

I like this paper :)


More like this

While cool, it would have been nice to see some control patients with diseases known to have a similar chronic involvement of microglial and macrophages within the brain paranchema (alzimers, CJD or parkinsons as examples). The production of retrovirus-like particles during chronic inflammatory diseases has been known of for decades (they were observed as far back as the 1960's, although no one knew what they were at the time) - and yes, that includes HERV-W in a chronic inflammatory neuronal disease (Alzhimers).

I'm not saying they deliberately biased their results, but they did manage to avoid "control" diseases previously shown in the literature to express HERV-W (and other HERV) transcripts in activated macrophages.

The immunologist in me is not too impressed with their disease model - aside from the assumptions they made of how MS is induced (they picked a controversial infection-induced model which is far from proven), their model is simply a conventional immune response with Env acting as a superangien in place of a conventional antigen. Not to mention you have to ignore a lot of what is known about the onset of MS to come up with this model (i.e. that the role of superantigens is largely theoretical with little supporting data) - and ignore the fact that no HERV genes have been associated with MS, despite numbers genetic, GWAS and other studies that should have picked up that sort of "universal" association.

The clinical trial is interesting; it may be a way of depleting lesion-specific microglia/macrophages. I don't know how well that'll impact disease, but it is far more specific than current therapies (which generally act to suppress the entirety of the immune system).

One concern I have as well is that all the positive staining in the immunohistology is from macrophages and microglial not from myelin tissue, per se. Those cells are part of the inflammatory response so how do you establish the link to autoimmune demyelination? Their hypothesis of herpetic virus crossing the BBB endothelium is not supported by repeated efforts to find evidence of those viruses in the brain of MS patients although the EBV work in that area is very contentious. If I saw expression of Env antigen in myelin tissue as opposed to inflammatory infiltrates, I'd feel better about the putative role of this ERV. One other thing, oligoclonal bands in MS are part of the differential diagnosis. These IgG band appear to be specific for, as yet, unidentified antigens. Has anyone determined if any of those antibodies are specific for the Env antigen? That work may be out there but I haven't seen it. One other thing that really doesn't seem to make sense. They saw higher levels of DNA in chronic progressive patients. These patients have much less active demyelination/neuroinflammation and the current belief is that the clinical deterioration of these patients is brought on by axonal atrophy and apoptosis caused by ineffective remyelination of the brain rather than ongoing inflammation. I would have expected exactly the opposite observation if this ERV plays a role in MS. I'd also like to see antibody levels to the Env antigen associated with disease activity in RR-MS. They point out that the patients were admitted to the hospitals for MS clinical activity. So the RR-MS patients were likely undergoing a relapse yet they had lower retroviral DNA. One other tidbit, this HERV has been in our genome for millions of years. Why was MS only first described in 1868? It's a pretty new disease when you look at that time frame. And how do you explain the extraordinary epidemiology that is observed with the disease with an endogenous retrovirus that everybody is carrying around in their genome? Something doesn't make sense there.

What you don't mention is that the Lombardi et al. serology assay would not react to any ERV and has only ever been known to react to MLV viruses. HERVs are known to reactive due to exogenous retroviral infections.

While I came to ERV for sanity in the elevatorgate ruckus, I am delighted by the real science discussion rather than cast off squid slime.

My wife has MS - so I follow these threads a bit more closely. And - after recently being transferred to Florida (state of ignorance ©)I am in a virtual blackout of real information. Thank you for keeping us up to date on what is going on in the real world.

And give Arnie some rubs for me!

By Bill the Cat (not verified) on 13 Apr 2012 #permalink

I have a problem with the bit where the authors talk of
"an increase of MSRV DNA copy number within PBMC for MS patients versus HC".

To me, this sounds as if MSRV in MS patients is actively replicating and re-intergrating into the genome. However, would this not lead to disruption, upregulation or downregulation of many other genes? Has this been observed?

The XMRV fiasco made me cautious and I realise that contamination can be everywhere. After all, Lombardi et al also reported a difference in XMRV prevalence between CFS patients and healthy controls. Therefore, I wonder whether DNA of the human investigators simply contaminated the PCR while they were looking for HERV DNA? Or am I too negative?

Serious comments (sorry, Gerwyn) would be very welcome!



While contamination is always a concern, this group did take some of the right steps to minimise that risk - i.e. samples were collected at multiple sites (so you shouldn't see site-specific contamination) and some/most/all (its not entirely clear) of the samples were tested at two independent facilities using RT-PCR devices from different companies. The chance of the same contamination entering just the MS samples at the different collection or testing sites is pretty remote; likewise, instrument error is also unlikely to replicate across two sites using two machines from different companies.

As for the copy number variations, there are a few possibilities. The first is that the MS patients inherited additional copies of HERV-W (which would be cool, as copy number variation could then potentially be used to ID people at risk at a young age). The second is that they are replicating in the cells - I don't know if this is possible for HERV-W though (I thought only HERV-K was replication-capable, but I very likely am wrong about that). The third possibility is that there are no extra copies, and instead the PCR reaction is picking up other HERV-W family members(HERV-W is a family of ERVs, as shown in figure 5, with only one member tentatively linked to MS; I think there is about 100 HERV-W members in total in our genome). Even that later case is potentially cool, if the cross-reacting HERV could be ID'd and shown to be associated with MS risk.

I still question the central assumption of this study - i.e. that HERV-W expression causes MS. Given what we know of ERV's and inflammatory diseases, it seems far more likely to me that the opposite is true - inflammation turns on the ERV, rather than vice-versa.

To me, this sounds as if MSRV in MS patients is actively replicating and re-intergrating into the genome. However, would this not lead to disruption, upregulation or downregulation of many other genes? Has this been observed?

Don't know, but they're arguing that the copy number is associated with ongoing RT activity, so it could be measuring an increase in newly reverse-transcribed DNA floating around in the cell - the actual rate of integration should be much lower. The authors also call for more investigation on how replicative these elements actually are:

"From a virological point of view, this requires further investigation on the more-or-less replicative activity of a peculiar MSRV/HERV-W element and on its eventual complementation or interference with other more-or-less defective HERV copies."

Interesting work. It is a pity that they did not include an ME/CFS group among the "other diseases" controls. If the ME/CFS community were not busy bickering and squabbling with each other and suffering a complete void of leadership, it may have been possible to have arranged for that sort of scientific cooperative effort.

"samples were collected at multiple sites (so you shouldn't see site-specific contamination) and some/most/all (its not entirely clear) of the samples were tested at two independent facilities using RT-PCR devices from different companies. The chance of the same contamination entering just the MS samples at the different collection or testing sites is pretty remote; likewise, instrument error is also unlikely to replicate across two sites using two machines from different companies."

Yes that is what they did in Lombardi et al. Well spotted.

@Levi: There are people like Nancy Klimas or Lucinda Bateman (to name the two people who are most likely least controversial while most qualified and versed in the disease) who would not hesitate to organize ME/CFS patients to participate in such a study, I'd reckon â no need for "central leadership".

@davey: Lomardi et al. was intentional fraud â so much in that study is smoke and mirrors. If they had written that the sun rises in the east every morning, I would check if it is actually true (and if they actually observed it themselves).

@ cynical1: I want to see were the research of Terry Wahls is going. If I had to bet money, I would bet on her â the epidemiology of MS would agree much more with her theory than with HERVs. We'll see, I hope.