Re: re: timeline and Ho 29 October 2004
Previous Rapid Response Next Rapid Response Top
Nicholas Bennett,
Infectious Disease Postdoc/Clinician
Department of Pediatrics, University Hospital, Syracuse NY

Send response to journal:
Re: Re: re: timeline and Ho

Mullis' objections are indeed well known. As the inventor of the PCR process he obviously has a great deal of kudos on things PCR, and his criticism (that a pseudo-mathematical amplification step cannot be used to measure anything) is a reasonable one! So why the rejection?

The point is, that quantitative PCR is not quite the same thing as what he is criticising. If one were merely running the reaction and then measuring the products, and then back-calculating to see how much was there originally, he would have a good case. There are however several steps one can take to ensure a PCR reaction is quantitative: from relatively simple serial dilution series, through branched-chain or multiplex PCRs, with internal and external controls of known standards. Then your result is simply fitted to a standard curve which controls for any variability. Off-hand I can't recall which method was used in this paper.

At the time, this was the best and most obvious way to show new production of virions. Later work has used by-products of the reverse- transcription reaction that the virus undergoes after cell entry to provide other measures of viral replication.

I don't think Ho looked at lymphoid tissue in that paper (he is listed by the way, as Senior Author) - the data was based on peripheral samples (and hence only looking at a small subset of the virus population). Pantaleo et al looked at lymphoid tissue [1] and saw much higher numbers and proportions of infected cells than in the periphery - the trouble is that taking lymph node biopsies isn't half as easy as a blood draw. The majority of data therefore comes from a sample set that is arguably not the best place to look for HIV.

From Pantaleo's work, I think he found HIV in around a quarter of lymphocytes in the lymph nodes. Far higher than the 1 in 40 CD4 cells in the periphery in some other studies. [2].

I don't think Ho stated he was looking at whole virions, although he does suggest that the levels are indicative of actual infectious virus (i.e. a high level by his measure would imply a high level of virus, but not necessarily a 1:1 relationship! It's far lower than that). However, as an RNA genome it would quickly degrade without some form of protection like a virion capsid and envelope, so it would be a reasonable assumption to state that the RNA levels were representative of whole virions, although statements about infectivity cannot be made without a culture comparison.

Nick Bennett njb35@cantab.net

1. Pantaleo et al Nature. 1993 Mar 25;362(6418):355-8. "HIV infection is active and progressive in lymphoid tissue during the clinically latent stage of disease."

2. Ho et al NEJM 1989 321:pp 1621-1625 "Quantitation of human immunodeficiency virus type 1 in the blood of infected persons"

Competing interests: None declared