EMSA with Proteins which bind non-specifically to DNA

[Adelaide_Science]

Hi All,

Whilst i am not a total novice to EMSA and DNA binding assays i have come up against a brick wall with some of my current proteins of interest. I am tired of banging my head, so i am appleaing to you more experienced users to try and help me out. I am a PhD student and as you can imagine time is of the essence.

I have used the Promega TnT to transcribe/translate my proteins of interest and am now looking to do EMSA with them. My proteins in other species bind to DNA non-specifically so i have got the X174 ss and ds DNA fragments from New England Biolabs to use to see the shift, as these fragments seem to be the concensus in non-specific EMSA.

I consducted an EMSA on 0.8% Agarose and stained with EtBr and just got this HUGE BRIGHT smear on the gel, obviuosly EtBr is staining up something in the master mix of the transcription/translation mix (the DNA and RNA in my thinking). So the normal EtBr staining is out for me!

I have labelled my protein with 35S methioine, so next i thought that i could test for the shift in the protein with and without the DNA. So i run the same amount of protein with no DNA and then a gradual increase of ssDNA on a 0.8% agarose gel and then looked via autoradiography of the shift. BUT my samples look like they are still in the wells of the gel even after 2 hours at 140V!!!!!!!! So obviously i couldn't see a shift in my protein running.

I am also thinking that the X174 DNA size is also close in size to my vector size, so if my proteins where swapping from the vector to the DNA fragment X174 then i wouldn't see it anyway? Though this time i was using the ssDNA X174 and whilst being 5386 bases, it runs at half the size cause it is single stranded, so in reality i should have seen the shift in the protein signal go down in the gel???

I have found papers showing that the TNT promega kit has been used for EMSA before, however they used short oligo nucleotides which their proteins specifically recognise as most people are working with Transcription Factors. I like to be different and work with recombination proteins!

My question to those of you in the know is that if i use an excess of shorter fragments (say aroung 30-40 bases) labelled with 32P most of my protein would bind to these and then i would be able to see a shift in the 32P DNA? Would i have to load in a unlabelled fragment of DNA (and larger) as well to soak up the other proteins in the TNT master mix which might DNA? Or would the vector and the RNA which is in the master mix (vector at 20ng/ul) be enough to mop up the other proteins and any of my proteins which aren't bound to the labelled DNA. Or do you thing that because my proteins do not bind specifically that any shift that i see with be due to any number of proteins in the master mix and that i could never get a conclusive result!Or could i get around that with using the master mix as a negative control?

Thanks for any help and any answers that you might be able to give me.

Adelaide_Science

[Adelaide_Science]

Hi again,

I forgot to aks about DTT in the reaction buffers which people use.

I use DTT often in protein gels to denature proteins, why are people using it in their EMSA buffers? Is it because it is at low concentrations and it just helps to get rid of some of the non-specific binding between protein-protein and protein-DNA? If i added it to my EMSA when using the protein label as a way of detecting the shift could i perhaps get a better result?

Also if the protein binding to the vector was an issue and therefore not seeing the shift; if i added a huge amount of a smaller fragment then X174/vector then really i should be able to see the protein shift from binding the larger fragments in the master mix to go to binding the smaller fragment. So would see the band of the protein go lower?

Cheers again

Adelaide_Science