This is actually pretty off-topic. However I'm not really sure of a good forum where you'd get worthwhile feedback (the biology stackexchange maybe, but that's pretty hit and miss), so I'll just go ahead and reply.

This is a pretty universal problem whenever you use a delta delta Ct method with qPCR. If you search through pubmed, you'll find papers describing searches for appropriate house keeping genes for normalization (e.g., this one on breast cancer). The long and short of it is that there's no universal gene that is appropriate in all tissues or all experimental contexts. 99.9% of the time, this fact is completely ignored and people will simply use GAPDH/ACTB/etc. and assume that their levels are stable enough. If these aren't stable enough in your context, then you pretty much have to search for an appropriate one yourself. One convenient way to survey common candidates is to just use a house keeping gene array, such as this one for rat from Qiagen.