Re: Virus built from scratch in two weeks

On Wed, 19 Nov 2003, Frank Martin wrote:

> Tim Fitzmaurice <[EMAIL PROTECTED]> wrote in message news:<[EMAIL PROTECTED]>...
>
> This means you can
> > start at a base point, bang out a few changes much more easily by
> > deliberately chanign the oligo synthesis and go for a live attenutated
> > vaccine.
>
> You can do that just as easily by regular molecular techniques.

Not really!  I spent the best part of 3 months trying to alter one base in
HIV using PCR-based mutagenesis.  I gave up modying the second strain and
cut/pasted the sequence across from the first into the right space.  The
mutagenesis was simply far far too fiddly and the virus kept looping huge
chunks of sequence out.  I think I finished up creating my own mutagenesis
kit from lab reagents, and left the fancy 160-pound box of goodies in the
freezer...

It takes about 24-48 hours to get a modified oligo sequence back from the
company - that's from putting the order into the office to getting an
envelope on the desk.  Changing sequence using that kind of technique is
far easier than mutagenesis, or so it seems to me.  Sure, you've got to
put it together afterwards, but you've got a head start it seems.  If you
can build a short sequence that includes some restriction sites there's no
need to build the entire 9-10kb genome, just a few hundred bases to
cut/paste into position.  YOU make it convenient, instead of relying on
nature to have done it for you.

> Speed of change then allows a faster response to new strains and
> > clades.
>
> That's all good and well, but explain to me how you can accurately
> predict against which epitope the immune system will best respond.
> That's almost impossible in a healthy person, let alone an HIV
> infected one.

<shrugs>  You can be far more specific with that than just chopping out
proteins, or bits of proteins.  Kill the LTR promoter in the 3' repeat, so
the virus cannot go beyond 1 replication cycle.  Remove the splice site so
no regulatory proteins are made (no Tat, no virus, pretty much).  Hell,
remove Tat and Rev entirely since they don't seem to be immune targets.
No worries about finding a convenient restriction site, just build the
sequence missing them out!  Why not just lose the RNA packaging signal, so
your little construct can make all the proteins and virus particles it
wants, but doesn't actually carry any genetic material to make more virus!
Lots of easy options with a powerful technique to create them.

FWIW

Bennett