antibodies binding DNA-collated replies

Fri Jul 9 11:07:00 EST 1993

> Dear All,

>    I recently ran a series of standard PAGE-retardation analyses using
> 20mer oligonucleotides labelled with 32P. The Oligo's contain the sequence
> for the binding of AP-1 (fos/jun). I retarded these oligo probes with various
> fibroblast cell nuclear extracts and got specific bands. To show the presence
> of specific transcription factors in this complex I added antibodies to c-fos
> and c-jun (ie in seperate reactions) and got what I thought were appropriate
> supershifted complexes. I have now had it suggested to me that these complexes
> may be due to the antibody serum binding the oligonucleotide probes directly.
> Does anyone know anything more about this phenomenum? It is new to me
> Thanks Graham Atherton
  i have some trouble trying to work out what you are saying, the complex
formation appears to be formed before you add the antisera, so all you are
saying is that in your detection system the antisera may be binding the
oligos?? that it what you are meaning??   please excuse my lack of
comprehension, its very early in the morning here.......yawn.
anyway, regardless of what you mean, antibodies to dna or rna are fauirly
unlikely i think, i have to work hard and be devious to get antibodies to
nucleic acids.  antibodies are the tools of my trade and i make lots of them.
the dna or rna needs to be haptenised to make antibodies against it, on its own
it is not very immunogenic, in fact it is very non-so.  for this reason i
consider it unlikely that this is what is binding in your experiment, but you
should obviously design a control to eliminate this possibility.

you could have some fun with just the antisera fragments, say the  Fab sections
or the like, and see what happens there, this could easily be your conytrol
experiment, and give usefull info

Rod Bonfiglioli, Waite Agricultural Research Institute,
University of Adelaide, South Australia.
e-mail rbonfigl at


I do not have any experience of my own.  I took a course taught by Phil
Leder 18 or 19 years ago in which he mentioned this phonomenon.  They
were beginning to work on immunoglobulin genes at the time, but I can't
remember if the original observation was while his group was working on
globin genes.

Good luck,

Jim Owens

  The phenomenon you are describing is real.  Some monoclonal antibodies 
isolated by hybridoma technology have been shown to have polyreactivity to
DNA, both double-stranded and single-stranded. I think this was first
demonstrated by Avrameas in 1982 and has been studied in immunology especially
for natural (auto)antibodies.  The problem you are refering to is that during
the development of antibodies in an immune reponse (like when one is
immunizing an animal with a specific antigen) there is also the production
of autoantibodies some of which are polyreactive.  These antibodies can
bind to DNA among other antigens like myosin and the Fc protion of IgG. 
So, not only do you have antibodies to c-fos but also weaker titers of
polyreactive antibodies.  Now these polyreactive antibodies shouldn't be a
problem because normally you would used the antisera at a titer far beyond
their reactivity range.  If you are concerned try and ELISA in which you
coat a microtiter plate with single-stranded DNA.  If your antisera binds
to single-stranded DNA at the dilution of your other assay then I would be
concerned about the results.  You can find ELISAs for DNA in Cairns et al.
J Clin Invest 74:880-887, 1984 and Cairns et al. J Immunol 135:3839-3844,
Good Luck,
denomme at FHS.McMaster.CA


Dear graham, I am not an expert in these matters, but two things come to mind

1. At low protein cocentrations non-specific binding of DNAby proteins can look 
remarkably like specific interactions particularly in tye abscence of 
competitor `DNA'.

2. DNA binding antibodies in human serum is rare and generally associated
with certain autoimmune diseases eg LUPUS. In fact lupus patients are 
excellent sources of anti-DNA antibody.


David Lonsdale    Lonsdale at


Thanks to everyone. This has been a big help and I have learned a lot.
I will now go and use some appropriate controls in my retards! 

More information about the Methods mailing list