Gel Shift Question

Song Tan tan at
Wed Mar 9 12:48:21 EST 1994

In article <genecutl-080394144743 at>
genecutl at (gc) writes:
>In article <smrodems-080394082728 at>,
>smrodems at (Steve Rodems) wrote:
>> I have started to do gel shift experiments and I have a question about
>> doing competition studies.  My probe is about 30bp.  Now, in a competition
>> study I use cold 30bp DNA as a specific competitor and linear BS/KS+ as
>> non-specific competitor (ie, should not compete off my specific complex). 
>> If I want 100-fold molar xs of competitor I can add 100ng cold 30bp DNA per
>> 1 ng 30bp probe.  My question is how much BS do I use?  100-fold molar xs
>> of BS actually has about 100x as many 30-bp sites since each molecule of BS
>> has about 100 30bp sites (that would actually be a 10,000 molar xs of 30-bp
>> sites if I'm thinking about this right).  So, what is the conventional way
>> to do this, molar xs of a whole piece of DNA or molar xs of available x-bp
>> binding sites?
>I think the best calculate the amounts of nonspecific competitor to use
>is to go by moles of nucleotide rather than moles of plasmid.  For example,
>1 mole of 30bp oligo = 30 moles of base pairs, 1 mole of 5kb plasmid =
>5000 moles of base pairs.

I agree with gc's response.  I would also add that for most binding sites and
vector length nonspecific competitor, the number of nonspecific binding sites
is essentially the number of bases in the nonspecific competitor.  For example,
if your binding site is 30 bp long and you use linearized Bluescript (about
3000 bp) as the nonspecific competitor, you will provide 3000 - 30 potential
nonspecific binding sites for your DNA-binding protein (each base of Bluescript
is the start of a new nonspecific binding site until you are within 30 bp from
the end of the vector).  3000 - 30 is not significantly different from 3000 for
most calculations of this kind.  Obviously, this explanation falls apart when
the binding sites are comparable to the size of the nonspecific competitor.  

As you've mused, the relevant parameter is the relative number of binding
sites.  This is particularly important to realize if one uses nonspecific
competitor of an comparatively undefined length, like sheared genomic DNA or
polyd(I-C).  One can always determine the concentration in bases by UV, but one
can't always determine the concentration of fragments in such situations. 

Hope this helps. 

Song Tan
Institute for Molecular Biology and Biophysics
ETH-Honggerberg (Swiss Federal Institute of Technology)
8093 Zurich, Switzerland
email:  tan at

More information about the Methods mailing list