Sequence non-specific DNA binding

Wolfgang Schechinger hubahopp at
Tue Aug 19 03:48:18 EST 2003

Dear Emir, 

something analogous to rnase protection assay comes into my mind: If your
protein binds ss DNA in a specific non-sequence specific way :-)), and you
digest away the non-bound part of the DNA, you should get some sort of
unique length DNA, you might be able to demonstrate these uniform size
fragments, when you label your DNA in a certain way (radioactivity might be
simplest). Maybe also non radiactive but sensitive detection like capillary
electrophoresis does the job.

In case of dsDNA, cloning the recovered blunt ended / blunted DNA and
determining the size and sequence of the inserts might give you lots of

If you have an idea of the size of the these fragments, by adding
synthetical oligos of different sizes each to your protein and measuring
the recovry, you could be able to determine the size dependency of the
effect and define a minimum length of ssDNA that will be bound.

Incubating the protein with a mix of oligos and doing MALDI/TOF or similar
procedures might be another possibility.
Fluorescent oligos and a suitable DNA sequencer also could do it.

Just some thoughts,


At 15:31 18.08.2003 -0500, you wrote:
>Hi guys,
>I was wondering if someone could shed some light on how to prove that a
>protein or a complex is binding single-stranded (ss) or double-stranded (ds)
>DNA and not just sticking to it when in pure form under given conditions. I
>did gel shifts with a protein I am studying right now and ssDNA under
>published conditions (Tris pH 8, Mg, ATP, ATP regeneration system).
>According to published data that are quite scarce, the protein should
>hydrolyse ATP in ssDNA-dependent fashion. Everybody seem to believe that
>ssDNA binding reaction thus required ATP hydrolysis. However, I found out
>that I can see the shift of almost the same intensity without addition of
>ATP and Mg (even in presence of 2mM EDTA and no Mg). Moreover, I am totally
>flabbergasted that DNA binding occurs even at 0-4C. Am I dealing with DNA
>sticking rather than DNA-binding? Is there a way to find out if DNA binding
>is real? The protein is not binding DNA in the sequence specific manner, so
>adding excess "other" DNA or poly dIdC would not compete off non-specific
>Any suggestions highly appreciated.


More information about the Methods mailing list