Thursday, August 12, 2004

Purifying HIV

Not too long ago I got in touch with Hanz Gelderblom, a renouned electron microscopist. The Perth Group have tried to use some of his review articles to state that HIV should be visible under EM of plasma samples. Since no such EM has been published, therefore HIV cannot exist (a ludicrous statement however you look at it). Hanz however has more specific things to say...

(Paraphrased at his request from the email)

Despite being successful at getting EMs of HIV from cultures, in parallel experiments using plasma samples he could not get enough virus. This was because they only recieved two samples of less than 0.5ml of blood from late-stage patients, when ideally it should have been far more from the seroconversion stage when virus titre is higher. He also attempted to get virus from Africa but couldn't get samples. Collaborations with groups from Sweden and Italy also failed because no samples of sufficient titres could be obtained. From experience, he thinks he could have been successful with particle concentrations of 10^10 per ml, about 10,000 times higher than what we here [at the BMJ Rapid responses] thought would suffice. Having tried for some time on this path, he gave up attempting to isolate HIV from blood about 10 years ago.

There are several logistical problems associated with EM of HIV from blood samples as well. The best way to ensure particle stability "in the field" is to fix the sample in higher then 0.5% Glutaraldehyde (GA), to prevent shedding of gp120 knobs. However, this level of GA prevents immunolabelling which he wanted to do to characterise virally-incorporated cellular MHC molecules, but could also be used to label virion proteins. As such there is almost a mutually exclusive situation where you can either keep the virus intact, or analyse it using immunoEM. He did attempt to look at virus from pelleted samples (rather than the 2-drop method mentioned in the previously quoted article) which would have massively concentrated the sample, but even this failed due to low titres of virus (see my earlier postings for calculations of virus needed to fill a 1mm cube).

He also is of the opinion (and a highly experienced opinion it must be) that any attempts to purify or enrich HIV from plasma will induce artifacts, such as changes in virion fine structure, shedding of gp120, osmotic damage etc. Obtaining particles with HIV morphology might therefore be a practical impossibility (much like trying to observe the position and speed of a subatomic particle simultaneously).

He does however admit that maybe plasma-derived HIV is more stable than that of cell-culture derived HIV, and gives the example of Avian Leukosis Virus as a comparison. We simply do not know, because no-one has been able to do it. He feels that direct isolation of HIV from blood samples might well be possible to do, and should be done since it would give a more realistic research material to study (not, notice, in order to prove its existence!).

I am deeply grateful for Dr Gelderblom for taking the time to write
to me, and for giving me permission to quote him online.


Calculations performed regarding the size of HIV particles:

4/3*pi*r^3 for a 120nm diameter particle, into a 1mm cube gives a little over 1 trillion particles. Average blood volume taken to be 5 liters.

The size of the particle is such that 200 AIDS patients with viral loads of 1 million per ml would need to have their entire circulating blood volume concentrated to produce enough virus to fill a 1mm cube (a grain of salt). Viruses are small, even if they can be prolific. The total RNA purified from that (bearing in mind that each virus contains 2 copies of the genome) is about 65 nanograms, or only just enough to visualise on a standard ethidium gel. Asking for an EM from a virus that hasn't been cultured or RNA from a virus that hasn't been amplified by PCR (or culture) is simply setting up a straw-man argument.


Post a Comment

<< Home