During any drug treatment, if any change in signaling is going on, the expression of phosphorylated molecules will be checked, according to my knowledge. But how is it possible?
It depends on your molecule of interest. Mass spec is often used to look at global changes in phospo pattern of molecules including proteins and lipids. there are also other redio-labeling approaches using an isotope of phosphorus.
For protein,s the easiest and most straight forward technique is immunoblotting using an antibody specific to phospho form of the protein.
You can also P32 label your experiment, IP your protein of interest, and use autoradiography for P32. The above suggestion is better, but my suggestion is necessary if there is no phospho-specific antibody. ;)
Another option is co-IP using an antibody to the protein of interest and a phospho-serine, threonine, or tyrosine antibody. Those antibodies are rather dirty, but necessary if P32 is a problem.
A final option is to check for a molecular weight change of your protein in a western blot with your favorite antibody. This is easier with a lower molecular weight protein, and usually not believed without a phospho-specific antibody. Or maybe a 2D gel would should the difference easily.
I agree that immunoblotting is the easiest method. If you run a two-color fluorescent blot, you should be able to separately detect the phospho-protein and the total level of your target protein (regardless of phosphorylation state). Those signals will overlap, but with the multiplex method you will very often be able to detect a small change in the gel mobility of the phosphoprotein, compared to unmodified protein (I've seen this many times). If needed, you can always use phosphatase treatment to confirm that what you're seeing is phosphorylation.