Infectious Disease Postdoc/Clinician
Department of Pediatrics, University Hospital, Syracuse NY
Send response to journal:
I did attempt to answer Mr Russell's points, but shall lay them out again.
There is no reason why an EM need be performed if other tests with appropriate controls will provide the same evidence. The EM picture will only show images of, hopefully, similarly shaped particles. The molecular techniques will then need to be employed (which De Harven didn't do in his early work, since the techniques didn't exist!) to actually prove the identity or uniqueness of the particles.
From what I can gather, there is no reason to conclude that De Harvens "isolation" of a virus was not a mixture of similar-looking virions! Where was the sequencing? There are after all more than 70 different murine endogenous viruses... These are rhetorical points of course, but no less valid.
One should also note that as likely as I am to speak the merits of molecular biology over EM, so too will a career electron microscopist tout the merits of EM over molecular biology... The truth, probably, is somewhere in between.
Since we cannot be sure that material which, molecularly, looks like being retroviral in nature is a pure virus, extra controls must be done. Attempting to detect the material in normal cells for example, or using additional purification methods to separate particles based on size as well as density. EM's from these additional purification steps DO exist . The biggest mountain of proof is that the proteins and genetic sequences found in these preps are not from the host cells. If you see a collection of proteins, of which 6 for example are unique to the putatively infected cultures, then it is a reasonable assumption to conclude that they may be viral proteins. If of course they look like antiviral chemokines or other immune proteins you're scuppered, but that's not the case with the HIV proteins. Additionally, EM of cultures with and without virus added will allow you to see whether or not a putative viral particle is an artifact or not. Molecular techniques can even label the vesicles to distinguish them from the virions, confirming the fact that they are different types of object .
There is only a superficial reason to assume that De Harvens technique will work on HIV, when it was developed for a different virus. The most glaring problem with it is that the cultures of Friend virus did not produce cellular debris - if they did then his system would not have worked! As such it is impossible to purify HIV by his method, in much the same way as it is impossible to sieve whole grains of wheat without first grinding them into flour - even though a sieve will work perfectly well on sand without further treatment. If you perform methods similar to De Harven's and then do EXTRA steps then the virus can be purified. My question therefore is, why should these results be ignored? Gelderblom has said that extremely high titres of HIV are required to get reasonable EMs, which reference  would attest to. Far higher than would be seen in an infected patient.
Mr Russell says that "indirect" virological markers will not suffice. With respect, there is nothing more direct than being able to detect, purify and fully characterise the genetic material of the virus. "Virions" are useless if they cannot replicate: the Dane particle of hepatitis B preparations is so unusual it has its own name, even though it IS the virus. The HIV genome has in fact been isolated not just from in vitro cultures, but also from infected patients .
Mr Russell also greatly misrepresents the practise of purifying HIV, by stating that current methods do not use ultracentrifugation, pelletting down of virus etc. Although he does not work in the field, I cannot see how he can miss these being clearly stated in the materials and methods sections of the papers he must have read. Molecular methods are usually only useful for detection, not purification and concentration (and ultimately, all a "purification" is is a relative concentration of one thing over its surroundings), and the older techniques still apply here: the only difference is in how the virus is detected, by EM or molecularly.
Additionally there is very clear reasons why the HIV virions can be labelled even within an EM field of mixed particles: because they look like retroviruses, with an envelope, spikes, a core and are of a certain size.
A dogmatic insistence of requiring a result that cannot be obtained due to technical limitations is poor science. Far preferable would be to try to understand why the alternatives were developed and what they can tell you.
As such, it seems logically appropriate to me that molecular techniques can be employed where others have failed. EM is not a gold standard, if anything should be it is the molecular cloning of a pathogens genome, since that is the blueprint of the entity.
More importantly perhaps I ask Mr Russell to look at figure 1A of reference  and tell me if that is to his satisfaction as regards purified virions? I note that they used a modified rapid-harvest sucrose- centrifugation protocol to keep the virus intact while removing contaminants. While it has its own set of interesting findings (a rather high proportion of multiple core-containing virions) the picture at least looks pretty, which is really all Mr Russell seems to want from EM.
I'm also waiting for Mr Russell to acknowledge that he was wrong when he said that no animal retrovirus was transmitted horizontally, and that he accepts that preferential male to female transmission is entirely normal for such a virus . Since a significant part of his argument against HIV being pathogenic apparently depends on this paradigm, one would expect it to be given due importance.
Nick Bennett firstname.lastname@example.org
1. Briggs et al. EMBO J. 2003 Apr 1;22(7):1707-15. "Structural organization of authentic, mature HIV-1 virions and cores."
2. Gluschankof et al. Virology. 1997 Mar 31;230(1):125-33. "Cell membrane vesicles are a major contaminant of gradient-enriched human immunodeficiency virus type-1 preparations."
3. Fang et al J Acquir Immune Defic Syndr Hum Retrovirol. 1996 Aug 1;12(4):352-7. "Molecular cloning of full-length HIV-1 genomes directly from plasma viral RNA."
4. Portis et al. J Virol. 1987 Apr;61(4):1037-44. "Horizontal transmission of murine retroviruses."
Competing interests: None declared