I have electroblotted on a pvdf membrane a 4-20 PAA-TGS gel run with samples 3 lanes containing DNA and a recombinant protein and 3 lanes of control containing only the protein. All the sample have the same concentration of the protein and of DNA.
I know that my protein is not stained by the Coomassie.
Once i have performed the ECL, I have the expected bands of my protein (in the low molecular weight region) in all the lanes (as I expected) but in those lanes where I added DNA there is a band in the very high molecular weight regions that appears as a ghost and the bands of the protein in this step have a reduction of intensity with the respect of the lanes where I did not loaded any DNA.
Furthermore these bands are colored by the coomassie stain.
What happened? Did the coomassie stain interact with DNA? Ir it is my protein that has interacted with the DNA and was stained by the coomassie?
Are you observing a consistent-weight ghost band present in all of your lanes?
So, you ran a native gel with your protein and a protein DNA mixture and you see a shift.
Do I get that right?
I'd like to understand whether it is a shift or not. i have only these data
A) did you run the DNA only control? (To answer the question in the title by yourself)
B) Did you do a gradient titration?
C) How does it look if you stain for DNA?
D) What happens if you boil the DNA-protein mix before loading?
E-Z) etc.
Stefan has some excellent suggestions above; you need to run the proper controls, especially the DNA-only control, before you can gain experimental confidence about what is going on with the high molecular weight bands you are observing.
The mechanism by which Coomassie blue stains proteins involves two forms of non-covalent bonding under typical conditions. 1) Van der Waals forces involving nonpolar amino acid residues. 2) Electrostatic interactions predominantly involving basic, positively charged amino acid residues; this is why proteins with predominantly acidic, negatively charged residues are stained more faintly by Coomassie.
DNA is negatively charged and does have hydrophobic regions, so I think it is possible that the faint heavy band could be your DNA without any protein bound to it. Of course, the other possibility, as hinted at by Stefan above, is that the protein interacted only weakly with the DNA and what you are observing is stained protein. Therefore, at this point, it is my firm opinion that you cannot properly interpret your result without further testing.
Thank you Daniel for the answer.
yes, the molecule within the coomassie stain could interact through these kind of interactions.
Even though I know that the coomassie blue molecule in particular is in the non-protonated form, so that it interacts basically by its negatively charged sulphonic moieties with the low pI proteins. This fact could explain why my protein, that has a quasi-neutral pI doesn't interact with the coomassie.
However the coomassie interacts also through non-polar interactions, because of its extended aromatic and planar system in the middle of the molecule.
but I know also that this interaction is typical when the coomassie is in the neutral form, that is the case of the coomassie form used in the bradford assay. In fact there are some works that report evidences about the interaction between coomassie and DNA, because of a shift in the absorbance peaks.
So In my opinion it's rather low the possibility of an interaction between the coomassie and the electroblotted DNA, also because the Coomassie should diffuse to the blotted dNA and intercalate between the bases by means of an EtBr-like mechanism, and in these conditions it seems to me quite unlikely.
So i'm wondering why in the western blot i see a dumped signal and in the blotted membrane in correspondence of that dumped bands i see a band coloured by the coomassie stain.
Silver ions are soft lewis acids, consequently it is expected that can interact not only with the soft moieties of the protein, but above all with the phosphate moieties of dan backbone. so silver stain can not discriminate between dna and proteins since i expect that can interact with both of them
Hi again Antonio, I share your hesitations about the hydrophobic interactions between Coomassie and the DNA being adequate for staining. However, it may not need to be an intercalation; hydrophobic surfaces are present along the grooves of the DNA helix, and in fact many DNA binding proteins target these regions. But of course, that is just a speculation. I still think that the DNA-only control would be very helpful for ruling out the DNA-dye interaction; especially since the bands are so faint, I am not altogether convinced that the dye is not interacting weakly with the DNA.
I thought of another possibility. What if the binding your protein to the DNA causes a conformational change in the protein that makes it more susceptible to staining by Coomassie? Except, if I am not mistaken by your description about the Western results, the antibody isn't picking up the protein in the high molecular weight bands, so that seems to rule that out... unless, of course, the antibody targets in or near the DNA binding region of the protein! Just a thought.
The only other thing I can think of at the moment is that some unknown "junk" protein, present at low abundance, could be binding your DNA with high affinity; how pure is your protein preparation? Can you confidently rule out possibility that an unknown protein might be the culprit behind your observations?
Anyway, thank you for the stimulating discussion, I agree that you have a truly mystifying set of observations. I sincerely hope you are able to figure it out!
Daniel thank you very much for your answer, I'll re-perform these experiments using these suggestions!
- Badges
- Science method