Infectious Disease Postdoc/Clinician
Department of Pediatrics, University Hospital, Syracuse NY
Send response to journal:
I note that the Perth Group fail to address any of the key points in my recent post (e.g. the cohort studies) nor do they refute that any of the mechanistic detals of HIV's immune-suppression are wrong. I therefore conclude that they accept that HIV is able to cause immune suppression and that HIV serostatus and HIV serostatus alone is predictive of AIDS. Of course my conclusions are likely to be wrong (about the Perth Group, that is).
They quote a series of caveats about CK that any houseman ought to know. This doesn't change a thing about the point I was making, which was that if you look hard enough (e.g. CK isoenzyme) you can come to a conclusion about the source. The same thing applies to RT activity - I ensured the reference explained this for the Perth Group, and having them quote it back to me, implying that I was somehow misrepresenting the contents, is laughable.
I'm astonished that they ask "If...Nicholas Bennett knows which of his cultures are infected and which are not, then why is he doing the test?" when they have made so many points in the past about controls! Unlike the Perth Group I do not make assumptions from my armchair about a state of affairs. When the experiment in question is looking at replication of HIV, one of the industry-standard surrogate markers of a viral culture (any virus) is a protein or enzyme assay, such as RT. Even in experiments looking at protein or RNA content, it is standard practise to ensure sufficient virus is present by performing an RT assay at some point: it takes half a day and can save you days of purification of an empty sample, not least allowing you to normalise different virus preps to the same RT value (e.g. for a prep with half the activity I will add twice as much volume - giving the same amount of virus).
Since I know which cultures are infected versus uninfected it therefore seems extraordinarily coincidental that only those that are infected show a rise in RT activity, despite the endogenous RT activity found in the cells (which the Perth Group would have us believe is sky- high - in practise it seems quite low). The fact that the uninfected cultures have a far lower level of RT activity proves that the endogenous activities are not significant compared to those in a real viral infection. RT assays also correlate with p24 antigen ELISA, which is an alternate method for quantifying HIV titre. It is all very well to argue the point that endogenous sources MIGHT be problematic, but in practise they are irrelevant - providing one performs the appropriate controls of course, and uses the correct method. Controls alone are sufficient to exclude cellular origins for the RT activity: I ask again, what explanation would the Perth Group offer me for an uninfected RT activity of 30 Units and an infected activity of 1000 Units?
The original methods used to characterise HIV RT (in the mid 1980's) did indeed pick up higher levels of endogenous activities: Montagnier isn't the only group and the fact that the Perth Group attempt to refute his data almost exclusively implies a personal vendetta. There are various means to remove or inhibit the cellular RT-like activity, as well the Perth Group know. In his "isolation" paper Montagnier performed these controls and others have done so also. The experimental evidence far outweighs any theorising from the sidelines.
Since the Perth Group have issued the invitation to refute them in formal print, I would be happy to do so. Not least because Montagnier's work contributes only to the molecular biology of HIV and not whether or not HIV causes AIDS, his "contribution" to that question ought to be relatively easy to refute. I found half a dozen glaring errors in the abstract alone. The rest of the paper appears to have been repeated ad nauseum on the Rapid Responses, and therefore the material to refute it is also here. Saves me a lot of time.
In his post of 8th Sept, Christopher Noble states:
"Nobody is saying that HIV infection is a necessary or sufficient cause of any of the AIDS indicator opportunistic infections or malignancies".
Which is true. The Perth Group twist this to mean:
"If AIDS can develop in the absence of "HIV", what are the other agents which cause AIDS"
Mr Noble did not say AIDS can exist in the absence of HIV, he said the AIDS-opportunistic infections can exist in the absence of HIV. By the same argument the Perth Group are saying that since KS can occur in the absence of anti-rejection therapy, the KS seen under anti-rejection therapy is not due to the therapy. Same principle - a wrong one.
HIV has not been the only factor considered as causing AIDS. Herpes viruses of various species, drugs and other lifestyle factors were all considered: and rejected. Only seropositivity to HIV was an independant risk factor for AIDS. Other factors modify that risk, of course: but then that is called Biology. This history is available for anyone with the urge to look.
As regards the formation of a viral structure, I refer the Perth Group to the oft-quoted paper from Richieri et al. which I emailed them on the 23rd July. There are of course plenty of other papers out there looking at membrane constituents, full-length genome cloning and protein content - some of which have been quoted here before. I suggest they perform the PubMed search themselves.
As regards the cellular poly(A) versus viral poly(A) RNA, I feel the urge to point out that "Poly(A)" RNA has very very very little, at all, whatsover, to do with retroviruses. In the very early days of virology it was peculiar that "poly(A) RNA" (meaning single-stranded RNA with a long strand of adenosines at the 3` terminus) was one of the characteristic features of the viral preps. The Poly(A) tail is in fact one of the anti- degradation and recognition features of messenger RNA (mRNA) used to produce proteins. The cells are FULL of poly(A) RNA. Even in a full- blown retroviral infection merely 1% of the cells poly(A) RNA will be retroviral, the rest will be the host mRNAs. For the non-virologists, 1% cellular RNA being of a viral origin is very high indeed. The fact that virions contain almost exclusively viral RNA is a testimony to the specificity of their packaging mechanisms.
The reason why retroviruses have a poly(A) tail is that they are transcribed by the cellular RNA polymerase II. They look and act exactly like cellular mRNAs, with the exception that certain viral genes or RNA elements prevent splicing from internal splice sites in some transcripts.
As such, to even ask why "cellular Poly(A)" RNAs are different from HIV RNA highlights the fact that the Perth Group are unaware of what exactly poly(A) RNA is.
And as to what HIV RNA has that the endogenous retroelement RNAs do not have...well. They have packaging signals, coding exons (Gag, Pol, Env), are associated with the appropriate enzymes etc. If by "the right conditions" the Perth Group mean hugely artificial situations, then ANYTHING can be made to integrate. This is the entire premise of the retrovirus gene-therapy approach: but the gene of interest still has to be packaged within a modified viral genome.
I clearly stated that my answers were a likelihood-based response. The chances of a cellular RNA containing and associating with all the necessary elements required to integrate are small: the chances of an exogenous RNA, introduced randomly into the cell, integrating are practically nil. Even the endogenous elements are largely defective and rarely move around the genome (by that I mean are transcribed and successfully re-integrate elsewere).
Polio is an RNA virus, and also has a Poly(A) tail to the genome. If the Perth group can find an example of polio genome being integrated into nuclear DNA I'd be hugely impressed.
I find it interesting that the Perth Group stated once that DNA polymerase alpha and gamma can appear like RT, and later said it was DNA polymerase beta and gamma. Either way, from the literature I've read it seems as if the activity in other DNA polymerase preps is likely due to contamination with the mitochondrial DNA pol III (pol gamma). Additionally, since the activity of the cellular DNA polymerases will be higher in the presence of certain conditions (e.g. Mn2+) whereas viral RT will be higher with different conditions (e.d. Mg2+) why can they not see that a simple pair-wise comparison will tell you the origin of a certain RT activity? The relative difference will be about 3-fold, according to the literature quoted by the Perth Group, in their Continuum article of Sept/Oct 1996. A decent delta for most people. I urge the Perth Group not to ask for the references, since I looked them up almost exactly 5 years ago!
I repeat, Montagnier stated he performed these comparisons: so why do the Perth Group persist in stating he did not?
Competing interests: None declared