I see. I don't know about toxicity because I worked with GFP variants and cells were fine. For photo bleaching it depends what type of FRET you are doing. If clover-mRuby2 is better that's great. I feel like pairs are less important. Important thing is how to interpret the FRET data.For example if you do live cell imaging:
1. FRET depends on conc.
2. How do you take care of stochastic FRET?
3. How do you determine the stoichiometry of the complex?
Another good pair that overcomes the photobleaching issues of Cameleon (CFP-YFP) is using mTeal-Cerulean. Yet, importantly, is the fact that blue fluorescent proteins (CFP, mTeal - even GFP I think) are chloride indicators, i.e., they report chloride levels intracellularly (they can be used to report on GABAergic activity).