Infectious Disease Postdoc/Clinician
Department of Pediatrics, University Hospital, Syracuse, NY
Send response to journal:
If Mr Russell was referring only to oncoretroviruses (as I well knew) then he should have said so. Such lack of clarity can only confuse the issue further. Since he admits that he cannot extrapolate the work of De Harven to other virus types, does he perhaps admit that to expect the same of lentiretroviruses is (somewhat) unreasonable?
Additionally, one wonders if Mr Russell is aware that the De Harven methodology, as I questioned previously , is not actually capable of isolating pure virus? As with many of the early "oncogenic" viruses, the virus was actually composed of a mixture of replication incompetant transforming virus(es), and transformation incompetant replicating virus . Perhaps some other form of isolation should be performed, like genetic cloning? Since we're on the topic of murine retroviruses, does Mr Russell accept that these entities exhibit sexual horizontal transmission with a preferential male to female spread, which he claims should not be possible for HIV? 
Prof Mullis' remark clearly states that "quantitative PCR" is his issue, rather than PCR detection per se. Since he did not invent quantitative PCR it seems a fair enough comment to make - his method of PCR was indeed largely non-quantitative unless you were to perform simple limiting dilution assays. Other workers have taken his original idea and drastically improved upon it, to the point where some of the latest quantitative assays do not actually use the PCR reaction at all [4, 5]! And as to who is in a better position to know X, Y or Z, I find it ironic that Mr Russell of all people holds this position, since he is clearly debating in a field of which he has no formal training whatsoever. I don't find it a problem to disagree with Prof Mullis any more than Mr Russell finds it a problem to disagree with Dr Bennett!
As I have stated here before , the amount of RNA in a virion is tiny - even 1 trillion virions would hardly contain enough RNA to be visible on an agarose gel without some form of amplification, and in order to get that many virions from a peripheral blood sample you would need the entire circulating volume from 200 people with viral loads of 1 million per ml! Hardly feasible. Culture and PCR on the other hand provide a standard scientific approach to amplifying what is already there to more manageable levels. Does Mr Russell have issue with using telescopes to amplify the sight of a distant star or planet? Why with PCR for HIV?
In addition, blood virus titre is not necessary to cause a disease, and I refer Mr Russell to the aforementioned viruses (Influenza, Rabies, Epstein Barr Virus, Human papilloma virus and Adenovirus) to ask if he can cite papers of significant peripheral viral load, sufficient to cause disease (no doubt by some arbitrary criteria of his own).
If I had access to an EM and a category 3 suite, suitable training in EM preparation, and of course the time, I would be happy to attempt the kind of experiments asked by Mr Russell. However, as Hans Gelderblom has said, virus titres are likely to be too low unless a modified protocol is used, but since isolation has been achieved for cultured HIV perhaps the methods employed there might be transferable (rapid harvest, anion exchange chromatography, Optiprep centrifugation etc).
When I'm in a position to perform such an experiment I will ensure it is at the top of my agenda.
Nick Bennett firstname.lastname@example.org
1. Nick Bennett, Rapid Response 20th Dec 2004
2. Ney and Andrea, BLOOD, 1 DECEMBER 2000 VOLUME 96, NUMBER 12 "Friend erythroleukemia revisited"
3. Portis et al, J Virol 1987 Dec:61(4):1037-1044. "Horizontal transmission of murine retroviruses."
4. Versant HIV-1 RNA 3.0 assay (bDNA)
5. NucliSens HIV-1 QT assay
6. Nick Bennett, Rapid Response 28 June 2004
Competing interests: None declared