Did Levy "isolate" HIV?

Todd Miller - Pharmacology tmiller at newssun.med.miami.edu
Fri Sep 8 10:14:50 EST 1995

Analysis of Jay Levy's paper "AIDS Retrovirus (ARV-2) Clone replicates in
Transfected Human and Animal Fibroblasts".
Nature 1984 312: 76-763.

There is no correlation between the data presented in the paper and the
title.  Perhaps it would be useful to first state what conditions should be
satisfied for claiming cloning of a virus and then analyse Levy's data.


1.    Have a particle which contains, among others, proteins and nucleic
      acids (RNA or DNA).

2.    Show that there is a direct relationship between the particle's
      nucleic  acids and proteins i.e. the proteins are coded by the
      nucleic acids (the viral genome).

3.    Introduce the viral genome (RNA or cDNA) into cells and show that the
      DNA (cDNA) is transcribed into RNA and the RNA is translated into
      proteins (transfect the cells).

4.    Show that the cells produce particles and that the particle's
      proteins are coded by the particle's nucleic acids.

5.    Show that the particle's nucleic acids and proteins are identical
      with those of the initial particle.

7.    Because all cells contain retroviral genomes, which under the right
      conditions may be expressed in culture, that is both the cells in the
      culture from which the original particle was obtained as well as the
      transfected cells may release identical retroviral particles even
      when there is no transfection, when one attempts to clone a
      retrovirus a control culture is of pivotal significance.

      There should be only one difference between the control and the cells
      transfected with the viral genome.  Instead of the viral genes one
      must use some other genes.  This is essential because the act of
      transfection and the condition used may lead to retroviral


-     Neither in the above paper nor in any of the references they cite for
      "ARV Isolation" are there any electron micrographs of viral

-     For transfection they used the cDNA of an approximately 9-Kb RNA
      which was selected from the culture medium of HVT-78 cells "infected
      with ARV", without ever presenting evidence that it belongs to a
      particle.  (This is also the case for all the other HIV genomes).

-     They have no evidence that transfection actually took place not even
      that the cDNA actually was incorporated in the cells.

-     There is no evidence that the "transfected" cells released any

The only evidence which Levy and his colleagues present as proof for "ARV
cloning" is:
(a)               detection in cultures of RT activity

(b)   detection in "Extracts of normal PMC", "infected" with "virus
      recovered from transfected" cells of proteins which reacted with sera
      from AIDS patients.  It is not stated what they mean by infection
      with "virus recovered from transfected" cells.  In all his other work
      on HIV, Levy, as well as all other HIV researchers, by infection
      means cultivation of cells with either supernatant from "infected"
      cultures, or the material from these cultures which in sucrose
      density gradients bands at 1.16 g/ml (without proof of the existence
      of particles).


(1)   RT activity is not specific to HIV or even to retroviruses.  In fact
      given the culture conditions they have used, it would be surprising
      not to detect RT activity irrespective of "infection with HIV".

(2)   The following proteins from the cellular "Extracts" were found to
      react with the sera from AIDS patients.  p16, p25, gp41, gp120 and
      gp160. =20

Neither in the above paper nor in any of his previous publications on HIV,
Levy (or any other HIV researchers) presented unambiguous evidence that the
above proteins are indeed coded by the 9-kD RNA he used for transfection.=
However, there is evidence that the proteins are non specific.

A few examples:


Although Levy did not find a gp41 in the "Extracts" from the normal "non
infected" cells, Montagnier did.  (Barr=82-Sinousi et al Science
1983;220:868-871).  The difference may be in the fact that apparently
Montagnier stimulated the non-infected cells but Levy did not.  According
to Gallo the molecular weight of the glycosylated protein corresponding to
the "HIV gene" which is supposed to code for gp41, "should be about 52.000
to 54.000 daltons", not 41.000 daltons.  (Medrow et al J Virol 1987;

gp120 and 160

There is evidence that the proteins of molecular weight 120-160KD which in
the WB react with sera from AIDS patients "represent oligomers of gp41",
that is they are not coded by a specific HIV genetic sequence.
(Pinter et al J. Vir. Meth. 1989; 63:2674-2679; Zallo-Pozner et al NEJM
1989; 320:1280-1281).


Monoclonal antibodies to p18 react with dendritic cells in lymphatic
tissues of a variety of patients with a number of non-AIDS related
diseases.   (Parravilini et al  AIDS 1988;  2:171-177).

The protein is not detected in normal non stimulated cells, but it is
detected in the stimulated "non HIV infected" normal cells.  (Stricker et
al Nature 1987 327:710-717).


The reaction of AIDS sera and of monoclonal antibodes to p24 which from
"infected" cultures bands at 1.16g/ml, is obiquitous.  The whole blood
cultures of 49/60 (82%) of "presumably uninfected but serologically
indeterminate" individual and 5/5 serogenative blood donors were found
positive for p24 by Schupbach et al (AIDS 1992 6:1545-1546).  [Schupbach is
the principle author of the third paper (the first paper ever on HIV WB),
in the series of 4 published by Gallo et al in 1984 in Science on HTLV-III

94% of sera from gay men with lymphodenopathy or AIDS react "with a 25KD
membrane antigen found in platelets from healthy donors and AIDS patients,
as well as a 25KD antigen found in green-monkey kidney cells, human skin
fibroblast, and herpes simplex cultured in monkey kidney cells".  (Styker
et al NEJM 1985; 313:1375-1380).


In the above paper by Levy et al there is no evidence for viral cloning or
even transfection of any genes, viral or non viral.


Now for question one.

"Why do AIDS patients tend to be pos for "HIV DNA/RNA"?

There is no evidence that this is the case.

1. By using Southern blotting neither Gallo nor anybody else
could prove that patients with AIDS are positive for "HIV
DNA/RNA". Even Gallo admitted this at the NIDA meeting in
Gaithersburg, Maryland, 24 May 1994. "I don't know if I made
this point clear, but I think that everybody here knows -- we
never found HIV DNA in the tumor cells of KS.  So this is not
directly transforming. And in fact we've never found HIV DNA
in T-cells, although we've only looked at a few". (This was
quoted in Continuum recently, see below).

2. Since 1987 the Southern blot test has been forgotten and
all the claims are based on PCR. However:

(a) in most if not all instances, researchers look for small
fragments of the "HIV gag and pol genes". The gag gene is
group specific, that is, it is shared by all human
retroviruses including endogenous retroviral genomes which are
said to be present in all of us. Positive results have been
found by many researchers including R. Weiss in non-AIDS
patients even by the use of the Southern blot. The same is the
case for the pol gene. (The retroviral pol gene is also
homologous with HBV pol gene. See Chang 1989 in our
Haemophilia paper and hepatitis B infection is not unknown in
AIDS patients).  As far back as 1985 Montagnier and his
colleagues, Simon Wain-Hobson and Mark Adizon, who performed
most of the HIV-1 genetic studies, wrote,  "We have also
compared the deduced amino-acid sequences of LAV proteins with
those of HTLV-1 and other retroviruses and find no significant
CONSERVED AMONG RETROVIRUSES".  (Nature 1985, 313:743). One
year earlier, 1984, Gallo's group reported that the "HIV
genome" hybridised with the "structural genes (gag, pol and
evn) of both HTLV-1 and HTLV-11" {Ayra, 1984}. Perhaps the best
indication that PCR is not specific is Gallo's admission in 1994 that,
"We've never found HIV DNA in T cells" (Continuum) by the use of a more
specific test, southern blot hybridisation.

(b) The specificity of the PCR has never been determined. The fact that
cells which do not contain nucleic acids and buffers give a positive PCR
indicate the many problems of PCR. Reference 150 in Bio/Technology is a
wonderful illustration of this fact as well and well worth doing the
numbers on the data. You'll find lots of WB pos and no PCR and vice versa.

Eleni Papadopulos-Eleopulos1  Valendar F.Turner2  John M.
Papadimitriou3  David Causer1
1Department of Medical Physics, 2Department of Emergency
Medicine, Royal Perth Hospital, Box X2213 Perth, Western
Australia 6009; 3Department of Pathology, University of
Western Australia.

Correspondence: Eleni Papadopulos-Eleopulos.
Voice int + 61 9 2243221 Fax int + 61 9 2243511
email <vturner at uniwa.uwa.edu.au>

More information about the Virology mailing list