I do believe you can compare if you normalize the membranes with the housekeeping protein. 

I am skeptical that this method will succeed in accurately measuring the % of phosphorylation. It is not because of the use of separate membranes for phosphoprotein and total protein determinations. It is because the different antibodies used for the two measurements may not have equal sensitivities. What I think can be done, however, is to demonstrate that the variation in some condition causes a change in the % of phosphorylation. For example, you might look at the effect of the concentration of a kinase inhibitor.

Hi Ask, if you have no qualms about working with 32P labeled cells, you can normalize the cellular protein concentrations that you are loading onto the gels (by e.g. BCA.).

You are now loading equal amounts of cellular proteins onto the gels. No housekeeping protein controls needed. Run the gel, blot it and look at the radioactivity on film

I didnt want to influence possible contributing answers, by sharing by own thoughts right away. But here we go. I was under the impression, that since there so many variables involved in WB, that comparing two different membranes would be impossible. Protein migration during SDS PAGE, evenness and efficiency of protein transfer from gel to membrane, and variable binding specificity of antibodies, would contribute to two membranes not being (possibly) identical. And then it would be difficult comparing phospho-protein from one membrane to total-protein on another membrane?

What do you think about this? I dont know if GAPDH solves this issue...

Lets say that protein transfer on the membrane with total protein, was much less effecient than the membrane with phospho-protein: Then you would see a much higher "fold" phosphorylation? But then again, the machine I use to develop the blots, has no standard of how long it exposes a blot, but just stops when it thinks the signal is of a particular intensity. So that would mean that you would never be able to "quantify", but only determine a ratio?

Hi Ask, do you want to publish clean results?

If you are thinking of running blots to show GAPDH, degree of phosphorylation, and total protein on three separate westerns, you will probably spend ~6-9  months to get everything right and splice the pictures together.

If you have the cells labeled with 32P and run equal amount of cell protein (I have done this before), then you can get the answers in one blot in 24 hrs. and ignore GAPDH because you are loading the same amount of cell protein

Expose the blot to film, develop it, digitize it and do a pixel count of the hot areas. You will not be able to say how much 32P is there, but you will be able to compare the intensity of labeling between proteins. 

Thank you for replying Steingrimur. I do not have the opportunity to with with that technique, and thus I have to suffice with WB. I am just pondering on the drawbacks of using two membranes, to look at degree of phosphorylation of the same protein, since some say its impossible, and others its fine.

Hi Ask, you could run one (or 2 gels) (15 well) with identical samples and cut them up after blotting and probe them.   

Hi Steingrimur, that is basically what I am doing. I run two gels, with identical samples, and blot them separately with their respective phospho-protein or total-protein antibody.

Hi Venge,

running separate blots is fine so long as you normalize to the same housekeeper.  Note that to get a precise measure, you will likely need multiple technical replicates (I typically run duplicates per biologic and subsequently average 3 biologics).  Note that in practice, this approach has limitations as the the different primary and secondary antiboi