Hi,

In my opinion, your method is acceptable, I've been using the same method for my proteins.  The only issue I guess you should have is the use of different primary or secondary antibodies, but since I guess you're not, I think it's a perfectly correct method.

For your information, make sure to have the bands perfectly resolved, i.e. no overlapping in the intensity profiles, though I guess you're already aware of this.

Good luck!

Instead of an answer, I have questions from your narration! If you are able to see your modified protein band separately, why all this calculation? or am I missing something? The formula you have put together will provide the ratio of your modified protein to the original protein. Within a gel this may be fine, but if you wish to compare between gels, then this ratio has to further calculated including, a loading control. One precaution I would suggest is, mutating the protein may lead to change in antigen presentation for western blots, hence immunoblot bands should be carefully analyzed. Good luck!

I also think method is ok but have the same concern /question as Muralidharan, does your antibody bind the native and modified protein with same affinity? 

Thank you very much for your answers!

@ Muralidharan and Stoyan:

The question about the affinity is indeed interesting. As far as I know from the datasheet the antibody is raised towards the C-terminus where the protein is getting modified. The trends I obtain are close to the results from using the antibody against this particular modification, however I am not sure where I can use separate antibodies to measure modification of the protein.

@Muralidharan

The reason I use the quantification is to show that different modification enzymes are adding the mark in different time points. 

There are two ways to solve your antibody issue; try using another antibody against different region of the protein, in conjunction with this antibody. The best would be to knockdown the protein of interest and use your modified protein to rescue it. This would be clean data about your modification on the time course you are studying.Good luck.

I recommend using two-color fluorescent Westerns for this type of experiment. You can multiplex the modification Ab with a general Ab against your unmodified target, if the primaries are from different species. Then do ratiometric analysis of the two signals. You can do this even if the two bands are not completely resolved. Ratiometric analysis will give you relative protein levels across that blot, and accuracy should be improved because you're comparing the same samples on the same blot. To compare separate blots, you would definitely need loading controls and replicates would also be important – this can get tricky.

If you multiplex, it's important to run controls with each Ab alone, to make sure there's no crossreactivity. You mentioned that the Ab was raised against the C terminus, which is also where the modification occurs. There could theoretically be steric hindrance - if you use two primaries simultaneously, binding of one might block/impede binding of the other (it happens, but isn’t common).

The bigger issue has already been raised: that the modification may alter the affinity of the primary Ab raised against the unmodified C-terminus. If that is happening, it would affect the outcome when you directly compare the two signals. Is this a monoclonal or polyclonal? A polyclonal is definitely better in this situation; since it binds multiple epitopes, it’s less likely to be disrupted by the modification (although it’s still possible). But in this situation, you would ideally want to use an antibody not raised against the C-terminus.

It’s good news that you are getting similar results using an Ab specific to the modification. I think it’s fine to use a separate Ab to measure the modification in this experiment, and do ratiometric comparison. Replicates are important to evaluate the error/variability of your method. You see papers reporting very small changes in abundance (1.2-fold and such) but I always wonder if the method has enough precision to make this a meaningful result. Westerns can have a lot of variability (CVs are often 25% or more).

How are you detecting your Westerns? If you're using film, I suggest running controls to examine your linear range. Film saturates incredibly quickly, and that’s a problem for densitometry. Some of the data is lost because of the narrow dynamic range – the density of moderate and strong bands starts to plateau long before the point of saturation and “blowout”. Moderate or dark bands can give you similar densitometry results even when protein levels are different, especially in longer exposures. It’s also very difficult to accurately compare bands across a wide range of intensities (faint and strong), because a single film exposure can’t capture all the bands within the linear range.

I would definitely recommend digital imaging for what you’re doing (unlike film, a digital imager will usually tell you if signals are saturated). If the changes you’re seeing are relatively small, a fluorescent Ab method may be a better choice than enzymatic ECL detection.

Dear Stephano,

Thank you for the comment! I have a quite significant shift between modified and non-modified protein, so resolving is not an issue. Of course I cannot exclude the possibility of  aggregates, hence I have a question: could it also be a significant issue in the case of low molecular weight proteins?

I agree that other methods will give a more accurate estimation, but the experiment I am intended to do is just a backup of the other data.

Kind regards,

Ekaterina

Dear Muralidharan,

Unfortunately the rescue experiments are impossible, as knockdown is very hard to establish and knockout is lethal. 

I will try to repeat the experiment using antibody against other regions of the protein.

Dear Amy,

Thank you for your detailed answer!

Unfortunately we don't have available antibodies for modified and non-modified protein, otherwise, that would be a very convenient method as there are people using this system in house (as far as I know). The antibody for the protein of interest is polyclonal. Considering I am interesting only in the modification on C-terminus it seems to work specifically, although of course it is worthy to check other antibodies.

For detection we use ECL and the pictures are taken by BioRad Chemidoc.