The Recombinant Origin of XMRV paper - why the Paprotka et al. paper failed to support the authors' conclusions

Paprotka, Coffin, Pathak et al. wrote the paper 'Recombinant Origin of the Retrovirus XMRV' (Paprotka, 2011, DOI: 10.1126/science.1205292).

This paper claimed that the XMRV variant of Human Gamma Retro Viruses (HGRV) was accidentally created in a lab when human prostate cancer cells were passaged through mice some time between 1993 to 1996. That XMRV did not exist before that time (1993) and that the original patient was not infected.

They surmise that the laboratory cell line thus created (22Rv1) then carried the XMRV virus to other labs, where it then contaminated human samples of blood or tissue, resulting in the findings in all the positive HGRV papers.

This is problematic, firstly because the WPI, NCI/Ruscetti, FDA/Lo and NIH/Alter labs have never had this 22Rv1 cell line in their labs. Secondly, just because a cell line is infected it does not mean patients cannot also be infected. Thirdly, it cannot account for other variants of HGRVs. And more importantly, this paper failed to include details on a third assay used in the study (RT-PCR) and failed to properly screen the cells to determine whether they contained the XMRV variant.

What follows is a description of how the authors used the three assays. To help the reader understand this section, an analogy has also been included at the bottom of the page.

ASSAYS USED IN PAPROTKA ET AL.

As you can see from the diagram below, cells were initially taken from a patient with prostate cancer in 1992. It has been hypothesised that this person must have been infected with the XMRV variant, because the 22Rv1 cells, represented by the 3rd jar from the left, are known to contain the virus.

The authors of Paprotka et al. would have you believe that the patient was not infected and that XMRV was created from two mouse viruses when tumour cells taken from that patient were transplanted into laboratory mice. This is called xenografting and was done repeatedly to the cell line between 1993-1996. After which the cells continued to be developed for research purposes (no xenografting involved) and eventually two sub-lines were created, 22Rv1 and CWR-R1. The latter is represented by the jar on the far right.

To support this argument, the authors of Paprotka et al. screened several generations of the entire cell line derived from the patient, including early cells that had been xenografted into mice and later cells that had been xenografted into mice. These are represented by the first two jars on the left. They state that the early xenografts do no contain XMRV, but the later ones do contain XMRV and that therefore XMRV was created during this transplantation into mice.

"We conclude that XMRV was generated as a result of a unique recombination event between two endogenous MLVs that took place around 1993–1996 in a nude mouse carrying the CWR22 PC xenograft." (Paprotka, 2011)

Click on image to enlarge

It is now know however that a third assay was used in the paper. Vinay Pathak in his presentation of this paper at CORI 2011, a conference on retroviruses, specifically stated that the later xenografts (2nd jar on the left) were screened with an RT-PCR assay. This assay was not included in the paper, was not used on any of the other cells, and we have no knowledge of its sensitivity (minimum amount of virus the assay can detect). Consequently, it is impossible to know the amount of XMRV that was found in the cells of the later xenografts.

It is also known from the paper, that these cells were not determined to be from the same patient as the other cells. This can be seen in the small grey box, labelled B, at the top of the image. The later xenograft's cells were given the following designations, 2152, 2524, 2272 and 2274, none of which were included in that experiment.

"We verified that the xenograft samples (736, 777, 9216R, 9218R, 8R and 8L) and the 22Rv1 or CWR-R1 cell lines were all derived from the same person by performing short tandem repeat (STR) analysis at 7 loci (Fig. 1B and fig. S1)" (Paprotka, 2011).

Without having assessed the later xenograft cells to see if they were from the same patient, it cannot be assumed that they were. Without this evidence it is impossible to argue that XMRV was create during xenografting between 1993 to 1996. Cross-contamination of cell lines is also a common problem, and again without this data we cannot be sure that the cells were not already contaminated from a separate source. So why were the later xenograft cells not included in this experiment?

Given these omissions, the paper cannot be said to have shown that the XMRV variant was created in this prostate cancer cell line during xenografting.

In addition, of the two assays actually named in the paper (qPCR and PCR), the qPCR was demonstrated to have at best a sensitivity of 1-3 copies per 100 cells and did detect XMRV in the early xenografts, but then the authors used a PCR assay, with no known sensitivity, to re-screen the early xenografts. It is this PCR assay they based their results upon.

This assay (the PCR assay represented by the blue line) was at most shown only capable of detecting as low as 2000 copies per 100 cells, but did not detect XMRV in the early xenografts, even though the qPCR assay had shown the level was at most 1-3 copies per 100 cells. Far below the proven ability of that PCR assay. The paper therefore also failed to provide evidence that the early xenografts were not infected and failed to present evidence that the patient was not infected.

As authors are obligated to provide data on all assays used to produce results published in scientific journals, here the authors failure to do so in this paper cannot be seen to be in the spirit of science and can only harm future progress in our understanding of human gammaretroviruses.

Finally, a number of laboratory mice were screened (no wild mice) to see if they contain viruses that may predate the XMRV variant. However, it is impossible to support any argument of contamination on this evidence as the mice used for creating the 22Rv1 cells are unknown, and no direct link can made to the cell line. The authors also failed to isolate from a single source one of the viruses (found in the authors chosen mice) that they believe was one of the two that combined to create XMRV (PreXMRV-1). Instead independent sections of virus were combined and called a whole virus. Thus, a hypothesis cannot be constructed as one of the two viruses has not been shown to exist.

SUMMARY

• We have four types of cells; Early Xenograft, Later Xenograft, 22rv1 cells and CWR-R1 cells

• A qPCR assay detected XMRV in the 22rv1 and CWR-R1 cells, but not in the Early xenograft cells. It was not used to examine the Later xenograft cells.

• A RT-PCR assay was used to examine these Later cells and found XMRV, but was not used to examine any of the other cells. This type of assay is also more sensitive than PCR alone.

• Details of the RT-PCR assay were not included in the Paprotka paper.

• No one knows where the Later xenograft cells came from or if they were contaminated.

• Paprotka et al. argue that XMRV was created by the recombination of two different gamma retroviruses when the Early xenograft cells were passed through specific mouse cells. They say the recombination event is so rare that it could only ever have happened once.

• There is no evidence that the mice tested in the paper were involved in creating the cell line originally, and the very existence of one of those mouse viruses they claim recombined to create XMRV, is unproven.

ANALOGY

The following is an analogy to help readers understand why the type of PCR used is important.

Imagine you are given two beakers, one red and one blue. Each beaker contains about an inch of what looks like sand. You are asked to find tiny bits of iron filings which might exist in one or both beakers. You are allowed to pour the sand out of the beaker onto a magnetic plate, shake off the sand, and use a magnifying glass to see if there are any tiny pieces of iron filings stuck to the plate. You are given a rack of magnifying glasses to choose from, because the filings are so small that they are right at the limit what you can see.

You pour out the sand from the red beaker and can see no filings at all. You do the same with the blue beaker and are able to see iron filings stuck to the plate. The problem is that you used different magnifying glasses to examine the contents of the red beaker and the blue beaker. You also find out later that the power of the magnetism in each plate was different, but you don't know which was the stronger.

Nevertheless, this paper assumes, to extent the analogy, that the blue beaker contained the iron filings but the red beaker did not.

The sensitivity of PCR, a type of test, is dependant upon the adjustment of a number of variables: magnesium, salt, buffers, annealing temperatures, and so on. By not determining the sensitivity of the PCR assay before its use the authors have rendered the interpretation of the results void.