electrophoretic mobility shift analysis with bacterial outer membrane

[biochem0123]

Hi,

I am doing an electrophoretic mobility shift analysis with the outer membrane protein of a bacteria and DNA molecules that potentially bind it. These DNA molecules are biotinylated and are detected by chemiluminiscence. The strange thing is I am visualizing a reduction in intensity of the band but do not really see a shift. If I increase the protein concentration, I do get a corresponding decrease in intensity, but have never been able to see a discernible shift. If I compete it with a non biotinylated DNA, the intensity remains the same as that similar to the unreacted DNA control. However I really would like to see a band 'shifted up". I have the following questions

1. Is there any precedence for this kind of a behavior to be decided as gel shift?

2. Can I interpret that the DNA molecule specifically binds the outer membrane preparation?

3. What can be done to observe the traditional shift like that involving a transcription factor?

4. Does anyone have similar experience with gel shifts with OMPs?

Thanks

biochem

[CharonY]

I need a bit clarification first. You have a purified OMP for the shift with the biotinylated DNA and you use (strept)avidin conjugate for Chemiluminescence?

1st question: is the OMP known to have a specific DNA binding site (I am a bit unfamiliar with that concept as I usually only do shifts with TFs).

 

And the following happened:

1) you increase OMP concentration, but not DNA -> intensity reduction

 

2) you add unlabeled but specific DNA (constant protein concentration)-> no intensity change

 

3) you add unspecific unlabeled DNA (control) with constant protein conc -> no intensity change

 

Did I follow you so far?

2nd question: do you only observe a single band in all EMSAs?

 

My quick'n dirty interpretation:

If I assume that no technical errors occured (e.g. during chemiluminescence detection) then it is no sufficient proof for DNA-protein interaction.

First, of course, because no shift happened (d'oh).

Second, because the control (adding specific unlabelled DNA) did not result in an intensity shift (displacement from regulator).

The strangest part is possibly the reduction of intensity by adding the protein. It is possible, that the protein in fact binds the DNA, however for some reasons the DNA-protein complex did not migrate into the gel. In that case an increase of prot would deplete the DNA, whereas increase of DNA wouldn't change much as you'd only see unbound DNA in the gel in the first place.

[Bluenoise]

CharonY has a good point. How heavy is this protein? Could it not just be getting stuck in the well?

 

Also what mechanism are you using for chemiluminesent detection?

There's a good possibility that binding of the protein to the DNA molecule is blocking the chemiluminescent signal. If you've constructed the potential binding site to close to the biotin linker you might have some issues with steric hinderance. If this is the case likley adding a longer spacer arm for the biotin linkage or additional nucleotides, maybe necessary.

 

You should stain with Etbr to see if there is maybe a ghost band that you can't see. Or if your lab is equiped radiolable the oligo, this can be done even with a 5' biotin as you can add additional nucleotides before the biotin modification.

I'm a bit confused as to why you would go through all the trouble of this chemiluminescent visualization.

[CharonY]

I assume it is one of the kits which utilizes biotin-(strept)avidin-conjugates for visualization of the DNA. I usually just use cy-labeled oligos for PCR (easier, more direct and non-radioactive, saves paper work).

EtBr staining is a good idea, though.

I was just thinking, the OMP does not happen to bind to biotin, does it?

[Bluenoise]

Oh no I would definatley not recomend radiolabling for PCR.

But in this instance where there might be a band that can't be visualized with the current method it does offer far better sensitivity than anything else. Plus basically nothing can interfere with its detection.

Personly I only ever radiolable when working with very small amounts or under conditions where I can't visualize any other way. Radiation is a pain in the ass.

 

I was just thinking, the OMP does not happen to bind to biotin, does it?

 

Now wouldn't that be ironic.