Non-isotopic EMSA-dirty uneven background - (Jul/23/2007 )

I bought a set of LightShift EMSA kit for my DNA-protein study.
Two problems raised:
1. I have no problem to see the gel shift in control experiments. But, the background is dirty and uneven every time.[attachment=3293:DSCN2707_edited.JPG]
2. The intensity of free probes in competitive experiments were weaker than "labeled probe only" and "labeled probe + NE". If competition do happen, the labeled probe should be released and showed the equal intensity as "labeled probe only", right?
The binding reactions were constructed as recommended by manufacturer. Nuclear extract was obtained by home-made extraction buffer and 9 micrograms (~1 microliter) was used in each reaction.

I think your background problem can be reduced by more wash steps and weaker exposure and maybe diluting the streptavidin HRP.
In the competition reaction with the unlabelled probe your shifted band should disappear but won't when an unrelated or mutated unlabelled probe is used.

Ceri

The wash was extended to 15 min/each for 5 times, exposure was just 2-5 s. I also extended the blocking to 30-60 min. No improvement. I never thought to change the concentrations of any reagents that vendor claiming "optimized". Maybe it's worth trying. Thanks!
Someone had told me never touch gel, membrane ... with gloves (even they are powder-free). I did and still failed.
The only condition that came out with clean background was reused membrane treated with 8M Guanidine HCl (pH1.5) for 10 min then went on directly for detection. Although the signals still there I could not take the data obtained by unusual treatment.
Suppliers of membrane are also a issue. I am using Immobolin N+ (Millipore). Someones told me MSI and S&S were fine but not Roche and Hybone N+.
The unlabeled probe is the same with labeled one excepting for 5' biotin labeling.

I definitely don't think it's your reaction conditions - if it were, the effects would be only in the 'lanes' - so I personally wouldn't spend a bunch of time mucking about with your reactions at this point. I would run only the controls until you get a good blot, too, so as not to waste your samples -

I think there's a problem with your detection. are your reagents old? is it chemi detection? if it is film, I would think your film was bad, or that (more likely) your developing reagents are bad...I also agree with Ceri in that it looks like over-exposure to HRP or one of the other detection reagents, if it's not an issue with development itself. have you ever tried further diluting your HRP? are you certain that the buffer you are using to dilute your detection reagents was correctly made?

good luck

The LightShift contains all the reagents I need excepting NE and probes. And it is a fresh one.
Detection seems ok, because I did a test by directly dot a drop of labeled probe on clean nylon membrane and went to the detection part. Signal and background are fine.

Last week, I finally got a clear and clean results with the following modification:
1. Use of PIERCE Biodyne B membrane (positively charged nylon membrane) instead of the use of Immobilon Ny+ (Millipore)
2. 1:1000 fold dilution of streptavidin-HRP rather than 1:300 as recommended in the manual
I do not know why, because of: 1) Immobilonbut Ny+ is positively charged Nylon; 2) 1:300 is "optimal' as addressed in manufacturer's instruction. However, result is pretty good, just like any picture in the textbook. No background detectable after 30 min exposure. Later I will try Hybond N+ (GE) and Biodyne B (Pall, Pall is the producer of PIERCE's membrane, same stuff lower cost)
By the way, I found the texture of PIERCE membrane is pretty smooth while the Immobilon Ny+ has "hand-made"-like pattern as you can see lots of fibers spreading on the surface.

I also use the same kit and have the same problem about the intensity of the free probe which is much weaker in competitive experiments than in the "free probe only" or "free probe + extract" lane. Did you manage to find an explanation about this ??

Unlabeled free probe (competitive probe) will run at the similar rate with labeled probe. So, when you do the transfer, those unlabeled probe will bind to the nylon membrane too. Since the binding capacity of nylon membrane is limited, some labeled probe can not bind to the membrane. I think this is why the signal of free probe reduced.