Post a new comment on this article
Post Your Discussion Comment
Please follow our guidelines for comments and review our competing interests policy. Comments that do not conform to our guidelines will be promptly removed and the user account disabled. The following must be avoided:
- Remarks that could be interpreted as allegations of misconduct
- Unsupported assertions or statements
- Inflammatory or insulting language
Why should this posting be reviewed?
See also Guidelines for Comments and Corrections.
Thank you for taking the time to flag this posting; we review flagged postings on a regular basis.close
Posted by DustyMiller on 06 Jan 2010 at 14:13 GMT
The results reported in this paper are basically uninterpretable. The authors state that they failed to detect XMRV in CFS but they have not provided data to verify the sensitivity of the PCR assay employed. Given that XMRV is likely to be present in only a small number of nucleated blood cells, the sensitivity of the PCR assay is critically important. The statement that they were able to amplify human beta-globin in all samples is not reassuring in this regard because 2 copies of the beta-globin gene are present in all nucleated blood cells.
Typically what one would do to validate the sensitivity of the PCR assay for XMRV would be to perform PCR amplifications of patient DNA samples with and without the addition of a few molecules of XMRV plasmid DNA. A positive result in the sample with added XMRV plasmid would prove that the assay was capable of detecting XMRV at this level of sensitivity. A negative result in the sample without added XMRV DNA would then be interpretable. Such data were not presented. Even more rigorous analysis would involve DNA purification from patient cells with and without the addition of small numbers of XMRV-infected cells. This would allow validation of assay sensitivity and confirm that the DNA extraction technique was not affecting the ability to detect XMRV sequences by PCR. Again, no such data were provided. The only data provided that the PCR assay for XMRV is capable of detecting anything are some faint bands resulting from amplification of an unknown amount of XMRV plasmid in Figs. 1 and 2. The authors do state at the beginning of the Results section that "both primer sets (XMRV, MLV) were able to amplify a single target copy added to the reaction", but whether the reaction also included patient DNA is not stated and no supporting data are shown.
Given the importance of the original finding by Lombardi et al. that XMRV might be involved in the pathogenesis of CFS, which as the current authors state "is an incapacitating disease affecting millions worldwide", follow-up studies deserve a high level of scientific rigor. The is especially the case for studies that appear to contradict the earlier study. Unfortunately the current study misses that standard.
We state that we were able to detect a single copy of XMRV in patient DNA. Hence, the only way in which we would fail to detect XMRV in a spiked sample is if there were inhibitors present in the DNA prep and we have controlled for this by amplifying the beta-globin gene.
However, the experiment you seek is not difficult to execute and, in the interests if scientific rigor, we shall carry it out asap and report back.
Thank you for your response to my comment, and I'm looking forward to seeing the new results.
In response to your first paragraph, it wasn't clear from your statement in the Results section of the paper that detection of single copy XMRV DNA was performed in the presence of patient DNA, only that a single copy could be detected in "the reaction". And again, your ability to amplify the relatively abundant beta-globin DNA is not the best way to show that you can amplify DNA present at much lower amounts.
One other technical point - do you shear the high-molecular-weight genomic DNA from the patients, for example, using a small-gauge syringe needle? This should be done to allow accurate dilutions and measurement of genomic DNA concentrations. Moderate shearing through a fine syringe needle should not affect your ability to amplify XMRV sequences.
I do not think that we do shear through a fine gauge needle, but I shall check with the two first authors who carried out the work and get back to you.
Please see response from Drs Kaye & Erlwein:
Since we did not prepare the DNA ourselves I cannot be certain that King’s did not shear the DNA but I would be very surprised if they did as this is not standard practice for either genetic analysis or viral detection in human DNA samples.
2. Amplification of beta-globin per se does not rule out the possibility of low-level PCR inhibition which could impact on the detection of low copy number targets such as XMRV provirus. With this in mind we looked for evidence of low-level PCR inhibition in a sub-set of samples using real-time SYBR-green PCR. We found no evidence of any such inhibition. The real-time PCR outputs showed identical slopes and identical plateau levels.
3. I am not aware of any evidence to suggest that DNA-shearing substantially alters the quantification of DNA by spectrophotometry nor am I aware of any evidence to suggest that shearing will alter the efficiency of PCR amplification. Certainly, routine shearing of samples does not appear in standard protocols for the quantification of HIV and HTLV provirus in infected patients.
RE: RE: RE: RE: RE: Results uninterpretable
DustyMiller replied to mmcclure on 08 Jan 2010 at 21:16 GMT
Regarding the need to shear the DNA - I have spoken with several investigators who use PCR to analyze genomic DNA from blood cells and they do not shear the DNA. One stated that the DNA is sheared during purification so there was no need for additional shearing.
My concern stemed from our work using very high-molecular-weight genomic DNA generated by spooling the DNA. This material is difficult to disperse in aqueous solutions and is difficult to pipette because it remains as a sticky blob unless sheared. Presumably you don't have this problem and do not need to shear your DNA.
RE: RE: RE: RE: RE: RE: Results uninterpretable
mmcclure replied to DustyMiller on 08 Jan 2010 at 22:12 GMT
Correct. We did not have a problem.
Results of spiking experiment should be available on Monday.
RE: RE: RE: RE: RE: RE: RE: Results uninterpretable
Re spiking experiment:
We randomly chose 22 of the CFS patient samples and spiked the DNA with 10 copies of XMRV plasmid DNA. The XMRV /MLVsequences were amplified in every case. Positive and negative controls worked beautifully. All conditions for the PCR were as described in the paper.
RE: RE: RE: RE: RE: RE: RE: RE: Results uninterpretable
In your previous responses, you state that "Hence, the only way in which we would fail to detect XMRV in a spiked sample is if there were inhibitors present in the DNA prep and we have controlled for this by amplifying the beta-globin gene", later adding "Since we did not prepare the DNA ourselves I cannot be certain that King’s did not shear the DNA but I would be very surprised if they did as this is not standard practice for either genetic analysis or viral detection in human DNA samples."
Since the object of this added experiment was to try and minimize the possibility of procedural error, shouldn't XMRV be added to the patient blood sample and then extracting the DNA using the same procedures which were used to prepare the samples tested in the study, possibly even by the same operators in the same lab using the same materials, instead of just adding XMRV to the already prepared DNA?
I do not have training in retrovirology however so please pardon me if I'm not following correctly.
Thank you for your time.
RE: RE: RE: RE: RE: RE: RE: RE: RE: Results uninterpretable
This was the best we could do, since we wre supplied with the DNA. However, I think the point is proved both with this experiment and the inclusion of the beta globin gene.
RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: Results uninterpretable
Thank you very much for your reply. There is one more question I would like to ask, if possible. I have seen several mentions of the rather weak signal from the positive controls. If the controls are so hard to read, then shouldn't the primers be adjusted(or changed, since different primers were used between this study and the Science paper) until a strong signal is received from the positive controls and then test the samples?
Again, thank you very much for your time, it is very much appreciated.
RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: RE: Results uninterpretable
PLOS_ONE_Group replied to JohnM on 20 Jan 2010 at 20:55 GMT
Due to a technical problem with our site, a comment that was left on this thread between 5 pm PST 1/17/10 and 3 am PST 1/20/10 has been lost.
We apologize for this situation, if you are the author of the missing comment we encourage you to re-post it