I keep samples for western blots at -20 and I have run samples after two to three months storage with good results.  I'm sure -80 is just as good or better.  I never boil my samples more than once, but I do incubate in a 37 degree bath for 5 minutes to thaw and do a quick vortex.  This is to make sure the SDS is in solution as it precipitates at 4 degrees.  I have heard of people boiling frozen samples again, however, so I can't guarantee that it wouldn't have an advantage for certain proteins, but for me it's not been necessary.

I often perform IP experiments where I elute my protein in 100 ul of 2x sample buffer.  I store the eluted sample buffer at -80 and have run my samples multiple times.  Everytime I defrost the sample, I reheat it at 70 degrees for 10 minutes and everything works well.  Like most things with Westerns, there are multiple protocols and theories on what to do and what not to do.  

Hey Anna, I agree with aforementioned suggestions that storing some part and reheating is the best way to go. However, I would like to caution that if the amount of your IP protein or its interacting partner is low, which is usually the case with endogenous (or tissue based) IP, then saving half your IP means you are reducing the amount of your interacting partner by the same fold as well. This might lead to some proteins giving a false negative result with your blot. Stripping and re-probing might be a relatively better option if you can get rid of your antibody with mild stripping agents. That way you wont loose a lot of your protein in stripping and have almost the entire amount in your lane. However, if you are going to do the same for overexpression samples, go for it!!

Good luck.