HIV sequence from virus

[Todd Miller - Pharmacology]

    Marnix Bosch asked me to read 2 back to back articles in 
_Nature_ 312: 757 and 760, 1984, in response to my question
about HIV sequence from particles as opposed to provirus.  
First, only one of these can be HIV since the first article 
describes a clone for LAV and the second a clone for ARV-2.
Was it LAV that later adopted the more provocative name, HIV?
Second, both clones that are characterized are from genomic
DNA libraries made from infected cells in culture.

    Is there some technical problem with the following protocol
designed to clone the sequence in particles?

1.  Purify the virus by banding on a sucrose density gradient.

2.  Treat the appropriate fraction (1.16 gm/ml?) with RNase and
    DNase.

3.  Disrupt the viral particles by say, phenol extraction
    followed by precipitation of the nucleic acid left in
    the aqueous phase.

4.  Make double stranded DNA by reverse transcription and 2nd
    strand reactions (oligo dT or random primers?)

5.  Clone into a plasmid vector.

6.  Screen as many clones as possible to make a restriction map
    of a circle with consensus LTR's.  At least see if all the
    clones can be explained as being part of such a molecule.

7.  Assemble a full length clone if necessary by cutting and
    pasting fragments from overlapping clones.

    Has this been done or does all the HIV sequence come from
proviral DNA?  Is this technically possible and doesn't it
represent a more rigorous proof of viral nucleic acid, especially
given the absence of EM's showing anything that looks like
virus at 1.16?  It bothers me a bit that these fractions are
not treated with RNase before cloning steps.  This could only
ensure that sequences cloned after this treatment were protected
at the time by being inside the particle.  This is a common strategy
in the isolation of bacteriophage DNA where properties of the
particles are first exploited.

    If anyone knows of any articles where the large molecular
weight DNA of the cell has not been part of the cloning procedure,
I would appreciate the reference.  I'm just a little skeptical
of this given that we know the genome can be very dynamic and
there is not a real effort to prove the sequences cloned are
indeed from viral particles in the supernatant.

Todd

PS--Don, did you figure out why I wanted to know now?