1 M urea is not as strong as 2% SDS, so I would not expect it to dissolve your pellet if 2% SDS did not. I personally do not like precipitating protein for the reason you state; it’s hard to get them back into solution. If possible, design your protocol to avoid precipitation. If not, have you tried sonicating the pellet (sonicating bath, or tip if you are desperate) to aid solubilisation? The choice here will depend on what you have available and the volumes you are working with. You can also apply gentle heat.

You need to remove free amines before you label with iTRAQ, this **may** be the reason people are desalting prior to labelling. It could be they have other interfering substances and want to remove them prior to labelling. It’s not required, but that depends on your protocol prior to labelling.

The Pierce spin columns should be fine for your clean-up. If you haven’t used them before, I suggest you test them on less important samples and make sure they work in your hands. The last thing you want to do is use something new with your precious samples.

I don't think 1M urea will work, b'coz for me at times 6M Urea even has not worked and had to add 20mM DTT along with it and gentle heat were given in that kind of situation; so as Peter said try avoiding Acetone precipitation; if you can and look for something like buffer exchange using cut off membranes.