Reader Comments

Post a new comment on this article

Patients infected with a retrovirus once thought to be XMRV

Posted by RRoberts on 01 Mar 2012 at 22:24 GMT


The title of the study could in itself be considered misleading, in that the retrovirus detected by Dr Mikovits and Dr Ruscetti was once thought to be XMRV, but the virus remains unsequenced. This fact will change the entire meaning of the study from the perspective of an independent reader.

The authors have not stated that these were not the samples used in the Lombardi et al. (2009) study. These samples of blood were drawn from 5 patients (4 positive, 1 negative) who had previously participated in the Dr Mikovits/Dr Ruscetti study. Only 5 patients from the 16 different participants were involved in the original study.

The multiple sources of contamination, which so concern the authors, were actually present in their labs during their experiments. The three sources of contamination, namely, mouse DNA, the 22Rv1 cell line and a subline of the 293T cell line, were not present in the labs of Dr Mikovits or Dr Ruscetti during the performance of the experiments involved in Lombardi et al. These samples are not the ones used in Lombardi et al. and hence the contamination found is likely to have been introduced by the authors themselves. Indeed the results of the supernatant analysis is readily explained by contamination from the 22Rv1 cell line present in the authors laboratory throughout the experiment.

It is very difficult to understand why the authors did not repeat the flow cytometry, used in Lombardi to detect the presence or absence of the monoclonal antibody to SFFV env in the patient plasma samples, which cannot be explained by contamination, or indeed why they simply did not retest the Lombardi samples, which have already tested negative using a validated test for mouse mitochondrial DNA.

The decision to carry out a different PCR assay than the one used in Lombardi is also puzzling, as is the choice to carry out PCR on plasma. No patient plasma samples were reported to be positive for murine-related virus (MRV) sequences in Lombardi et al.

X-SGA analysis of the proviral sequences detected in the plasma RNA taken from the 4 CFS patients, that had previously tested positive in Lombardi, showed that all had a 9 nucleotide deletion in the glycogag leader sequence, which according to the authors involved, is unknown in any endogenous mouse virus. Therefore, their conclusion that they had detected a mouse ERV is clearly erroneous and mystifying.

In the course of this study the authors also chose to sequence the env region from only one patients plasma sample. The potential that the env sequences detected in this patients sample could be contamination from the mouse DNA of the mouse cell line involved in the experiment is enormous. This view is supported by the fact that Dr Mikovits and Dr Ruscetti could not detect any env sequences whatsoever and the env sequences reported by Silverman were that of the VP62 plasmid, which he accidentally contaminated his samples with.

This record of accidental contamination further reinforces the view that these env sequences were the result of accidental contamination by the authors themselves. The Lombardi samples have already being examined by Mr Switzers (CDC) mitochondrial DNA assay and found to be negative. This should have been made clear by these authors.

The mitochondrial DNA assays and IAP assays used in this study are unvalidated. The products were not sequenced. This is a glaring omission, as it is not known whether the IAP assay could detect human endogenous retroviruses, which are virtually homologous with mouse IAPs. HERV RNA production is greatly upregulated in people with immune dysfunction and would be in the presence of an exogenous retrovirus. Hence this would explain the positive findings compared to controls. The purported mitochondrial cytochrome c oxidase assay is also novel and its ability to differentiate between mouse cytochrome c oxidase DNA and human cytochrome c oxidase DNA has not been established. COX2 is upregulated in people with an activated immune system and hence the DNA in people with CFS is unmethylated and hence easier to detect with PCR than it would be in controls. The authors could have confirmed their suspicions or otherwise by simply sequencing their product. It must also be emphasized, that this testing took place after the status of the patients were known, leading to the real possibility that the samples were handled differently.

A rigorous search reveals that the 293T cell line can only be infected by VP62/XMRV by deliberate transfection and produces such a low titre of virus that it can only infect the LNCaP cell line used in Lombardi by co-culture. This cell line was not present in the Dr Mikovits and Dr Ruscetti experiments, and even if it had been, it cannot fairly be described as a potential source of contamination. The sequences recovered from the 293T cell line showed up to a 12-fold sequence variation compared to the sequences found in 22Rv1. The authors conclude that this must be due to evolution within the 293T cell line. This is despite the fact that an exhaustive literature search undertaken by an independent team reveals no evidence that such an evolution within a cell line is possible (12). Indeed why is this sequence diversity not seen in the 22Rv1 cell line? This sequence diversity argues against the authors preferred hypothesis, that the ultimate source of all xenotropic sequences in the human population is the 22Rv1 cell line. If the authors want to prove the existence of contamination in the Lombardi studies, then we would suggest they examine the samples used in Lombardi et al. rather than rely on assumption and innuendo. The suggestion the authors are making is that these samples were contaminated by cell line DNA in the NCI. There is however a more likely explanation, namely, contamination by 22Rv1 cell DNA present in the authors own laboratory.

• No samples were from Lombardi et al. These were fresh samples.

• None of the culture samples were from the Lombardi cohort

• Only 1 of the 5 negative plasma samples were from the Lombardi cohort.


Two sets of samples were tested in this paper. Blinded plasma samples and non-blinded culture samples.

The first thing to note is Table 1, which lists the plasma samples that were tested by the authors from 9 people. 4 were CFS patients and 5 healthy controls. Of those, all 4 CFS patients were tested in Lombardi et al., but only 1 of the 5 healthy controls. The “a” next to the numbers in the first column indicates which theses are. CFS patients are X2, X3, X5 and X8. X6 is the only healthy control from Lombardi et al. These samples were all sent blinded to the HIV Drug Resistance Program (DRP) lab and were never tested by Dr Ruscetti or Dr Mikovits.

Table 2 lists culture samples from a variety of patients. None of these were from Lombardi et al. as they had undergone different experiments prior to testing in this study. These samples were not sent blinded to the DRP lab who performed the testing.


The results from plasma testing on the 4 CFS patients and 1 healthy control that participated in Lombardi et al. is concordant with the results from that study.

All 4 CFS patients (X2, X3, X5, X8) tested positive for an MLV type virus. The 1 healthy control (X6) was negative. Only two of the other 4 healthy controls tested positive for XMRV. Interestingly only the four CFS patients that tested positive for an MLV type virus were positive for the unvalidated COX2 and IAP assays that were intended for use as mouse contamination assays. We will return to the mouse contamination assays later in this response.

“The Taqman probe used for detection of amplified products was designed to span the signature 9–24 nt deletion in the XMRV gag leader absent from all endogenous MLV sequences (with the exception of PreXMRV-2)”

According to the authors, a 9-nucleotide deletion in the glycogag is absent from all endogenous MLV sequences except in the case of the virus nicknamed PreXMRV-2. However, the MLV sequences detected in this study in patients with CFS all contain the 9-nucleotide deletion in the glycogag (Fig 1C). The following statement is therefore incorrect. 

“These data show that the amplification profiles from the CFS patient samples mimic those of MLV and not XMRV.”

Hereafter the MLV type sequences found in CFS patients will now be correctly referred to as MRVs. These MRVs are much closer to the VP62 sequence than the representative mouse ERV sequence, but they are not VP62, because they do not have a 24-nucleotide deletion in the glycogag, which VP62 does. As they are not ERV sequences, it is quite wrong to use the description that they mimic ERVs. They show greater sequence homology to VP62/XMRV in the glycogag region than they do to the representative mouse ERV. Hence if they mimic anything, they more closely mimic VP62 than the ERV. However, it must be said that mimic used in this context by the authors is a highly inappropriate description.

“X-SGS was also performed on the env gene from patient X8 and compared to single-genome env sequences obtained from mouse cells (Figure 3). The majority of sequences in patient X8 contained large deletions, as seen in the highlighter plot of the alignment of Figure 3b ( sequence/HIGHLIGHT/highlighter_top.html). Similarly large env deletions were also present in the endogenous MLVs amplified by X-SGS from mouse genomic DNA, again confirming that mouse genomic DNA was the source of MLV sequences in the plasma samples from the CFS patients (Figure 3b)."

This plasma sample from the Lombardi CFS patient designated X8 is the only plasma sample that was tested using the env X-SGS assay, only the gag X-SGS assay was used on all 9 plasma samples. Fig 3. clearly shows sequence diversity between the ERV env sequences compared to the mouse DNA. In 3A, one full-length env sequence (X8env-5) from patient X8 can be seen clustering near to modified polytropic viruses, and two full-length env sequences (X8env-4, X8env-10) can be seen clustering near to polytropic viruses. In figure 3B again genetic diversity of the sequences is evident. The authors make the following statement regarding the sequence alignment of these ERV env sequences in Fig 3B.

“Similarly large env deletions were also present in the endogenous MLVs amplified by X-SGS from mouse genomic DNA, again confirming that mouse genomic DNA was the source of MLV sequences in the plasma samples from the CFS patients (Figure 3b).”

Thus, in patient X8 (CFS), the gag sequences contain deletions not known in any endogenous mouse ERV, but env sequences very similar to those found in the mouse cell line used in the experiments. The simplest way to reconcile these observations is that the authors contaminated this PCR with the DNA of the mouse cell line used as a positive control. This begs the question, why did they not sequence the env region of the other three CFS patient plasma samples?

In Figure 2, what is not immediately clear to the reader is that the authors have grouped identical sequences of endogenous MLVs extracted from the C57Bl6 genome sequence so that they appear as one. For instance, “Mpmv3 (10)” is in fact 10 endogenous MLV sequences. The scale on this Figure (100nt) is also not the same as that used on the other figures, so horizontally the sequences look closer together than they would if using the scale from Figure 2 and 3 (10nt).

The taqman probe is digested from the 5 prime end, during extension of the primer. There should be no partial fluorescence, either the probe is bound or not, unless the assay is not specific. The authors are claiming that the low level of max fluorescence means it is the ERVs amplifying – meaning that the reaction kinetics are different when using two targets. There is no data to support this claim and any calculation of copy number based on this inefficient assay is going to be unreliable. If the authors were confident that the low levels of maximum fluoresence indicated the amplification of ERVs and not exogenous MLVs, it is difficult to understand the need to sequence. The authors should have emphasised that an alternative explanation for low maximum fluoresence levels is non-specific binding.


All 9 supernatant samples were tested with the env X-SGS assay, but only three, 1 to 3, were tested with the gag X-SGS assay.

“XMRV, but not MLV, sequences were detected in all 9 culture supernatants. As shown in Figures 4 and 5 and Table 3, all of the XMRV sequences detected in the supernatants were nearly indistinguishable from XMRV sequences found in well-characterized XMRV-infected cell lines. “

In Figure 4A and 4B, genetic diversity of these MRV sequences can be seen. These results can be readily explained by contamination arising from the 22Rv1 cell line present in the authors’ laboratory, but not present at the NCI. The authors also have the problem of explaining the following.

“XMRV produced by 22Rv1 cells shows very low diversity (Figure 4, 5, Table 3), whereas the XMRV sequences from 293T-XMRV cells and co-culture supernatants were 2-16-fold more diverse (Table 3), consistent with acquisition of a few mutations during rounds of virus replication in cell culture that occurred during the co-culture procedure that included 2–10 passages.”

Dr Coffin claims that XMRV was born during the creation of the 22Rv1 cell line. How then is it possible for 293T cells to produce 2-16-fold more XMRV than the 22Rv1 cell line? This is evidence that XMRV was not created during the passage of these cells through mice sometime between 1992 to 1996.

Considering that the full details of the western blot, that determined this subline of 293T cells to perhaps be infected with XMRV or an MLV, are not available for analysis, it is unsafe to assume the 293T cells were infected before testing with the X-SGS assay.

“, although the 5 plasma samples from healthy controls were free of mouse DNA, two of them contained contaminating XMRV nucleic acid – most likely plasmid or a PCR amplified DNA product – as shown by the PCR amplification profiles and confirmed by sequencing of the amplified product.“

The implication here is that the samples were contaminated at the NCI, but the laboratory of the authors contained a far more likely source of the contamination, namely the 22Rv1 cell line. This argument is supported by the fact that the sequences isolated had less sequence diversity than would be expected if they originated from the 293T cell line.

Research reveals that the 293T cell can indeed be transfected with the infectious VP62 clone (1,2,3), but there is no evidence that these cells can be contaminated other than by transfection. The cell line produces a very low number of virions (4) and can only act as a source of VP62/XMRV virions by co-culture with LNCaP cells. There is no evidence that virions within 293T cells can act as a source of contamination as far as LNCaP cells are concerned. The authors suggested that the 293T cell line was present in the NCI and the WPI where the experiments were carried out. This is false. They suggested that this cell line could act as a source of contamination in spite of evidence to the contrary, which is misleading.

The IAP and mitochondrial assays used here are unvalidated, both in terms of specificity and sensitivity. The samples used in Lombardi were subjected to the mitochondrial DNA assay created by Mr Switzer (CDC) and were found to be clear. It is therefore puzzling why the authors would chose to use an unvalidated assay, without ascertaining its ability to differentiate human and mouse COX2. One would expect COX2 transcription to be upregulated and copy number to be elevated in ME/cfs patients compared to controls (5,6) and that would make the DNA more amenable by detection with PCR (7). There are HERVs that exist in the human genome whose sequences are virtually identical to mouse IAP sequences (8) and the RNA of these HERVs would be at a much higher concentration in ME/cfs patients than controls (9,10,11). Even if one ignores the high probability of false positives with an IAP assay, the decision not to sequence the products of these products, but to merely assume their nature, appears worthy of censure.

Finally, if the authors suspected plasmid contamination of the two healthy controls, why did they not run an assay that could detect plasmid contamination? The most likely source of the contamination in these two samples is the 22Rv1 cell line present in the authors’ labs at the time the experiments were conducted.

“We handled all cell culture supernatants in an area designated for cell culture and not in clean areas designated for processing of patient samples.”


Only one patient (CFS) was included in testing for both plasma (X5) and culture supernatant (PID 5). In Table 3 the diversity of the env sequences from the culture supernatants was compared to the virus in the 22Rv1 and 293T supernatant. Patient 5 (X5, PID5) has 0.05% intra-patient diversity and 0.02% distance from 22Rv1 consensus sequence. However, only 2 sequences were amplified for the env region of this patient (Figure 4). To compare other sequences had anywhere from 7 to 12 sequences amplified in Figure 4. Gag sequences from this patient were also well provided for (Fig 2 and Fig 5). Why were no more env sequences obtained from patient X5/PID5 so that the comparison in Table 3 would be more accurate?


It is difficult to reconcile the work of these authors with the work of Hue et al. (12). The latter authors found that the genetic pairwise distance between the different proviruses detected in the 22Rv1 cell line were greater than the genetic pairwise distance between the clones isolated from patients. Yet these authors found virtually no genetic variability between the sequences they isolated. Something is clearly amiss. The precise route taken by the authors to determine the so-called “consensus sequence” of 22Rv1/XMRV has not been revealed. If the provirus sequences within the 293T cells have up to 12 times the genetic diversity of the proviruses found within the 22Rv1 cell lines as claimed by these authors, then it is difficult to conclude that the proviruses within the 22Rv1 are the only source of XMRV infection.

However, the findings of Hué et al. (12) clearly show that contamination cannot be assessed by PCR testing for mouse DNA alone, since several human cell lines harbor xenotropic MLVs that are closely related to XMRV.

“Next, Hué et al. PCR-amplified, cloned and sequenced XMRV gag, pol and env segments from 22Rv1 prostate carcinoma cells, an immortalized line known to harbor multiple integrated copies of the virus. Remarkably, the 22Rv1 sequences displayed average pairwise genetic distances that equaled or exceeded those of previously-published XMRV sequences from prostate cancer [1] and CSF patients [3], despite the fact that these patients were from epidemiologically unlinked cohorts. In addition, phylogenetic analyses of the 22Rv1 and patient-derived XMRV sequences strongly suggest that the patient sequences obtained to date [1,3] originated from one or more XMRV proviruses present in the 22Rv1 cell line [11].” (13)


• The analytical and clinical specificity and sensitivity of this X-SCA assay using these primers, cycling reagents and cycling conditions is unknown. This is not the same assay that demonstrated an ability to detect 22Rv1 XMRV in the blood of recently inoculated pigtail macaques.

• This is problematic, as the specificity of primer and probe design determines the efficiency and accuracy of quantitative real-time PCR.

• The authors state that the probe purportedly spans a region unique to VP62/XMRV and S-162/XMRV because no endogenous MuLV has a deletion in this area. This is not true as a simple BLAST search reveals other candidates with a deletion sequence in the glycogag. (14,15)

• The authors have not demonstrated that their assay cannot amplify human endogenous gammaretroviruses. The expression of HERV RNA is greatly increased in people suffering with an exogenous viral infection (10,11).

• This combined with non-specific binding of the probes and primers would account for the results observed in this different X-SCA assay.

• The authors had knowledge of the testing history of the samples before sequencing and testing for mouse contamination and had mouse DNA in their laboratory and 22Rv1 cell line at the time of testing.

• There is no data to support the claim that the combination of cycling reagents, cycling conditions and primer combinations used in Lombardi et al. can amplify endogenous mouse MLV sequences and no data to support the claim that this X-SCA assay can detect VP62 like sequences in the blood of people with ME/cfs.


• 22Rv1 cells – Human prostate cancer cells

• TA3.Cyc-T1 – mouse mammary cancer cells

• 293T – Kidney cancer cells

The 22Rv1 cells are contaminated with XMRV. This is a potential source of XMRV contamination that the HIV Drug Resistance Program (DRP) introduced into the environment used for the experiments.

TA3.Cyc-T1 mouse cells are derived from strain A mice mammary cancer. This is a potential source of mouse contamination that the HIV Drug Resistance Program (DRP) introduced into the environment used for the experiments.

“A subline of 293T cells was obtained from the LBS for use as an uninfected control, but was determined by Western blot to be infected with MLV or XMRV, was confirmed to be XMRV by sequence analysis, and is referred to as 293T-XMRV. Culture supernatant and DNA from these cells were subjected to X-SGS for comparison to virus reported to be isolated from patient plasma and tissues. We handled all cell culture supernatants in an area designated for cell culture and not in clean areas designated for processing of patient samples.”

Which subline of 293T cells this refers to is not included in the paper. This is potentially very important, as some cell lines of 293T cells are infected with MLVs (16) and 293T cells have been involved in XMRV-VP62 transfection experiments since 2006 (17). Furthermore, it is not known if the LBS were aware that these cells were to be used as a negative control. The western blot that determined this subline of 293T cells to harbor either an MLV or XMRV is also not included in the paper for analysis. Therefore we cannot rule out the possibility that the western blot cross-reacted to something other than an MLV or MRV that would produce a false positive when this test was run. Subsequent testing on the subline of 293T cells was performed with the X-SGS assay. It is at this stage that XMRV, MLV or MRV contamination may have occurred.

All three of these cell lines, which the authors blame as a source of contamination, were in the authors labs at the time of the experiments and hence are a very plausible source of the contamination found in the authors samples.

These very same cell lines were not present at any time in the labs of Dr Ruscetti or Dr Mikovits during the Lombardi et al. study.

IAP and COX2 tests

“We applied two previously developed assays to detect mouse DNA. The first was adapted from methods developed by W. Switzer, Centers for Disease Control and Prevention, Atlanta, GA, to measure the level of mouse mitochondrial DNA by detecting the cytochrome C oxidase subunit 2 (COX2) gene [25] (Appendix S3), and the second was adapted from O. Cingöz, Tufts University, Boston, MA, to detect intracisternal A particle (IAP) sequences [14].”

Both of these assays are adaptations of the original assays that have still not been validated. As explained above, both assays could be reacting to things present in humans and it is also very likely that mouse contamination occurred during testing in this study.


In Table 1 the authors show no mouse contamination data or X-SGS data for plasma sample X9. However, only the X-SGS result for X9, which is recorded as NTC, is said to have been, “Not tested”, according to the footnote below Table 1. As the COX2 and IAP mouse contamination tests for X9 are recorded as NT, without a “c”, an explanation for this abbreviation needs to be provided by the authors. If these tests were positive for this negative sample (both XMRV and MLV) then it would indicate these unvalidated mouse contamination tests produce false positives.


As further study is required into the association of MRVs with human disease, it is disappointing that no GenBank accession numbers are provided in the paper for the new sequences. We hope this oversight will be rectified soon so that fellow scientists can study the results from this paper.


• The authors state that LNCaP cells were co-cultured by LBS.

“LNCaP cells were co-cultured by LBS with plasma, PBMCs or tissues (collected from bone marrow biopsies)…“

What is the abbreviation of LBS in this instance? Was this co-culturing of LNCaP cells done by Dr Ruscetti?

• The co-culturing done here by LBS is not said to include “B cells” or “prostate tissue”. Therefore, where was sample 2 and sample 7 of Table 2 obtained from?

• The authors state that Virus-positive supernatants were used to infect human foreskin fibroblasts (HFFs).

“Virus-positive supernatants were subsequently used to infect human foreskin fibroblasts (HFFs).“

Were these cells previously positive for an MLV-related virus? Who determined these samples were positive? Was this experiment done in the DRP lab or any lab other than Dr Ruscetti’s? Were the HFF supernatants obtained from the same person as sample 2 and sample 7 in Table 2?

• Who provided the samples of HHF supernatants to the DRP?

• The paper designates samples of HHF supernatants “DRP ID 1-9 (Table 2)”. However, later in the text there are samples coded “DRP PID”, not ID.

“…a subset of three supernatants (PID 1, 2, 3).“

The samples in Table 1 are also called “DRP identifier (PID)”. The samples in Table 2 are called “DRP identifier”. The samples in Table 3 are called “PID”. Are all the samples called DRP PID in the remaining text the same samples as in Table 2? What is the difference, if any, between the two sets of codes “PID” and “ID”?

• In the footnote to Table 2 “b” is said to designate:

“b The indicated sources were inoculated onto LNCaP (8–12 passages) and used to infect HFF cells, which were then grown in culture for 2–10 passages, after which the supernatants were subjected to X-SGS.”

Does this mean that only those samples in Table 2 that were cultured 8-12 passages were those that were inoculated into LNCaP cells? Specifically, only sample 2, 5 and 8? If yes, were the other samples directly inoculated into HFF cells?

• One patient who had virus isolation performed is said to have had a plasma sample taken in 2010. They are coded X5 and DRP PID 5. Is this patient in Table 2 listed as “5”?

• In the paper these two statements using the gag X-SGS assay appear to contradict one another.

“X-SGS of env was performed on all 9 culture supernatants and on gag from a subset of three supernatants (PID 1, 2, 3). “

“X-SGS of gag was also performed using supernatants of cultures treated with samples obtained from PIDs 1–3 and 5. “

Were three supernatants (PID 1, 2, 3) treated with a mixture of samples from PIDs 1-3 and 5?


The evidence in this paper indicates that although contamination of these samples probably did occur from multiple sources present in the DRP lab during testing, this was not the source of the MRV sequences found to be infecting the four CFS patients. The authors should have at least considered this a more parsimonious explanation for the inconsistent results and should have therefore retested fresh samples using the same assays without contaminated cells or mouse products present in the lab. The simplest way of confirming the presence or absence of contamination in the Lombardi samples is to test the Lombardi samples. It is difficult to understand why this simple orthodox scientific protocol was not used.

• MRV sequences found in patients is not found in any infected cell line or mouse ERV.

• Lombardi samples independently tested negative for mouse contamination.

• This study used unvalidated mouse contamination assays.

• The experiment needs to be repeated by an independent team of researches testing the actual samples used in Lombardi et al. and ensuring that multiple sources of contamination are not present in the study, unlike the situation in the study critiqued here.

• The HFF cells used here need to be screened for the presence of IAP particles.

• The cycling reagents, cycling conditions and primer combinations used in Lombardi et al. would allow repetition of the methods used and eliminate a major confounding variable.

• The performance of any novel real-time assay can then be objectively compared to the performance of the assay, which appears to be the most successful in Lombardi et al., and their specificities and sensitivities compared.

• Testing the samples for mouse contamination should be the first step to avoid the very real possibility in this experiment that samples were handled differently because the testing history was known.

• Two real-time assays should be used targeting a sequence unique to the VP62 strain of XMRV and using a set of primers to allow for sequence variability. This would allow the performance of the novel X-SCA assay used here to be compared to a real-time assay, which has demonstrated the ability to detect VP62 like sequences in people shown to be infected using IHC.

• The flow cytometry assay used to detect monoclonal antibodies to SFFV env needs to be repeated.


1) Paprotka et al. Inhibition of Xenotropic Murine Leukemia Virus-Related Virus by APOBEC3 Proteins and Antiviral Drugs. 2010. Jounal of Virology. DOI: 10.1128/ JVI.00134-10

2) Sakuma et al. No evidence of XMRV in prostate cancer cohorts in the Midwestern United States. 2011. Retrovirology. DOI:10.1186/1742-4690-8-23


3) Groom et al. Susceptibility of xenotropic murine leukemia virus-related virus (XMRV) to retroviral restriction factors. 2010. PNAS. DOI: 10.1073/pnas.0913650107

4) Rodriguez et al. Xenotropic Murine Leukemia Virus-Related Virus Establishes an Efficient Spreading Infection and Exhibits Enhanced Transcriptional Activity in Prostate Carcinoma Cells. 2010. Journal of Virology. DOI: 10.1128/JVI.01969-09.

5) Mauro et al. NF-κB controls energy homeostasis and metabolic adaptation by upregulating mitochondrial respiration. 2011. Nature Cell Biology. DOI:10.1038/ncb2324

6) Lee et al. Mitochondrial biogenesis and mitochondrial DNA maintenance of mammalian cells under oxidative stress. 2005. International Journal of Biochemistry. DOI:10.1016/j.biocel.2004.09.010 http://www.ncbi.nlm.nih.g...

7) Lee et al. Increase of mitochondria and mitochondrial DNA in response to oxidative stress in human cells. Biochemistry Journal. 2000. http://www.ncbi.nlm.nih.g...

8) Patience et al. Packaging of Endogenous Retroviral Sequences in Retroviral Vectors Produced by Murine and Human Packaging Cells. 1998. Journal of Virology. vol. 72 no. 4 2671-2676

9) Toufaily et al. Activation of LTRs from Different Human Endogenous Retrovirus (HERV) Families by the HTLV-1 Tax Protein and T-Cell Activators. 2011. Viruses. DOI:10.3390/v3112146 http://www.ncbi.nlm.nih.g...

10) Nellåker et al. Transactivation of elements in the human endogenous retrovirus W family by viral infection. 2006. Retrovirology. DOI:10.1186/1742-4690-3-44 http://www.biomedcentral....

11) Perron et al. The human endogenous retrovirus link between genes and environment in multiple sclerosis and in multifactorial diseases associating neuroinflammation. 2010. Clinical reviews in allergy immunology. DOI: 10.1007/s12016-009-8170-x. http://www.ncbi.nlm.nih.g...

12) Hué et al. Disease-associated XMRV sequences are consistent with laboratory contamination. Retrovirology. 2010. DOI:10.1186/1742-4690-7-111. http://www.retrovirology....

13) Smith RA. Contamination of clinical specimens with MLV-encoding nucleic acids: implications for XMRV and other candidate human retroviruses. 2010. Retroviriology. DOI:10.1186/1742-4690-7-112. http://www.ncbi.nlm.nih.g... Smith 2010

14) BLAST search patient X2 and X3 sequence in Figure 1C : “XMRV complete proviral genome, isolate S-162” (FR872816.1) http://www.ncbi.nlm.nih.g...

15) BLAST search patient X5 and X8 sequence in Figure 1C: “Mus musculus mobilized endogenous polytropic provirus clone 51 truncated gag-pol polyprotein (gag) and envelope protein (env) genes, complete cds” (FJ544578.1) http://www.ncbi.nlm.nih.g...

16) Andrawiss et al. Murine Leukemia Virus Particle Assembly Quantitated by Fluorescence Microscopy: Role of Gag-Gag Interactions and Membrane Association. 2003. Journal of Virology. DOI: 10.1128/JVI.77.21.11651-11660.

17) Urisman et al. Identification of a Novel Gammaretrovirus in Prostate Tumors of Patients Homozygous for R462Q RNASEL Variant. 2006. PLoS Pathogens. DOI:10.1371/journal.ppat.0020025.


Ronald Roberts, ME Advocacy Association

No competing interests declared.