It all depends on what purpose the GFP serves in your CoIP. Not being too familiar with the technical aspects of CoIP, and what it is you are CoIP'ing, it's a hard question to answer.
Generally speaking, the main differences between fluorescent proteins that are of interest to most people are (i) wavelength of absorption/emission, (ii) maturation time, (iii) oligomeric status, and (iv) brightness. Here is a nice list of FP properties: http://www.fpvis.org/FP.html . Sequences between FP's are often similar enough that antibodies cross react but there are no guarantees.
There are plenty of plasmids on Addgene with GFP genes. You can always get one and clone it out and fuse it in frame with your protein of interest and insert into your vector. Or just order a custom Gblock from IDT with protein of interest + GFP already in frame, and then just insert it into your vector. If there is any reason to doubt compatibility at all, I see no need to compromise a lengthy and expensive series of experiments (CoIP's) simply to avoid PCR's to make an established construct. I expect my students to be wizards at cloning to avoid having to make such decisions.
Edited by labtastic, 22 November 2017 - 01:13 PM.