Why GFP/YFP in neurons is lost at cytometer after mouse brain dissociation? - (Jun/10/2015 )

Hi everyone! I'm trying to isolate parvalbumin-positive neurons of hippocampus and cortex in transgenic mice with YFP expressed only in parvalbumin neurons (Cre/Loxp method). I want to extract RNA from these cells in order to make transcriptomics.
My protocol is based on decapited mice, extract brain, disintegratation with papain solution, 70 micrometer filtration, a percoll gradient to eliminate debris (I tried a lot of percentages: 30%, 40%, 90%..) and finally I directly go to FACS with all brain cells washed after percoll to eliminate any residue. But I had never seen any YFP-positive cells: control mouse with no YFP expression and mutant mouse with YFP expression show the same pattern, no fluorescence.

I also tried to use other mice expressing GFP under another promoters (GAD1-EGFP,...) but it was the same: no fluorescent cells found.

Does anybody knows why GFP/YFP could be lost in my preparation? Because I see papers where they use this type of transgenic mice and have no problems with seeing the GFP at cytometer and they don't do anything different from my protocol. HELP please! I've been trying different protocols a long time and nothing changes.

Did you once monitor GFP/YFP expression before cell separation? Bringing the whole acute brain to fluorescence with appropriate UV radiation, or better checking expression in an acute brain slice in fluorescence or LASER scan microscopy?

Another hint is that GFP is sensitive to oxidative stress and daylight which means that you may have expression but it´s lost during stressful preparation regarding GFP/YFP. One given stressor in your preapartion is Percoll.