EMSAs - Some probe questions - (Aug/24/2007 )

This is my first go at EMSAs and I am having trouble with my probes. I ordered 5' biotin labeled 24mers from a local company and annealed them in an old PCR machine by incubating at 95C for 5 minutes and then reducing to 4C by 1C/min. I annealed them at 1pmol/ul in STE (50mM NaCl). I used the Pierce Lightshift kit and got blazing bands with their controls, but my probe was really weak, probably 50-100x less bright than the controls. I re-annealed and now I don't see anything at all from my probes, but the controls are still bright. I am running 20fmol of probe and only running the dye (loaded only in the Pierce control samples) 2/3 of the way down a 6% gel. I have 10% glycerol in my binding buffer so the samples are sinking in the wells.

I talked to the oligo company and they swear they get 95% biotinylation, but I am skeptical of the probes. I am planning to re-order them from Sigma-Genosys, but I wanted to see if there may be other suggestions before doing this.

Also, I am basing my oligos around a sequence that has only been used a couple of times in EMSAs and never in my system (Xenopus). One of the papers used a 22mer with 7 bases on either side of the 8mer consensus sequence and the other used a 30mer. The flanking sequence in both of the papers and in my system is different. Am I better off designing a longer probe? I've read that probes are anywhere from 20-50bp and sometimes much longer, but I'm not sure how to design an appropriate probe. My probe contains no hairpins and has a GC content of 35%.

Thanks,
Kurt

you can do a labelling efficiency check on your probes, but it's probably cheaper/easier just to order a set from somewhere else.

couple things to think about:
there might be something interfering with binding?
your TF may be present in super-low concentration - is the smear at the bottom of the gel nice and thick? (would you please post a pic?) do you have stimulated/unstim controls for your system?
you may have bad oligos - you may need to tweak your binding buffer.

you may have oligos that are not what your TF is recognizing. longer oligos ... hmmm catch 22 there. you may get better binding (particularly if your TF binds to some secondary or tertiary structure of the sequence? it might need more bases) or you may just be picking up binding of a different TF to an adjoining site. sometimes the DNA binding site is ~4-6bp long; too much unnecessary sequence and you'll pick up all sorts of background. one thing to do - to test your oligo specificity - is to see if you can get purified TF and run the reaction with a small amount, to make sure the shift you're seeing is correct. even so, I would try to prove it with a supershift if at all possible, once you get the system ironed out.

good luck