Question about gel shift assay

Dirk Seegert seegert at
Thu Jun 20 05:03:37 EST 1996

SAI-Jinling Xie <xiej at IRCM.UMontreal.CA> wrote:

>Hi, there,

>     We've been doing gel shift assay. But we always got non-specific 
>shifted band. We used more than 220bp PCR products as probes and crude 
>cell extract from E.coli for shifting. A company recommends very short 
>probes. I wonder if our probes are not short enough or some reasons else. 
>And does impurity of the protein cause this problem? Do I have to purify the 
>protein? If somebody happens to have the similar experience, would you 
>please reply to this account? I would highly appreicite it. Thanks a lot 
>in advance.

>Jinling Xie 

We use probes of about 25 to 30 bp. Using longer probes may result in
conformational changes so that your desired binding site isn't
availible for proteins. Moreover your flanking sequences could work
like competitor DNA reducing  your specific signals. If you're
searching for any binding site in your DNA probe I would recommend to
cut your probe into smaller fragments and use this as competitors.
Another way is to perform  normal binding reactions (about 30 min, RT)
followed by exonuclease III digest and running the probe  on a
denaturating urea gel.
Hope this helps. Don't hesitate contacting me for further questions or

Dirk Seegert, Ph.D.
Fraunhofer Institute Hannover
Nikolai-Fuchs-Straße 1
30625 Hannover
e-mail: seegert at
Tel.: +49 511 5350 255

More information about the Methods mailing list