We want to inform the community about recent observations regarding photoconversion of GFP. Kevin O’Connell has observed photoconversion of two different reporters that contain codon-optimized superfolder GFP (constructed by Andy Golden). No fluorescence is observed with initial 561 nm excitation (mCherry filter) but, after visualization @ 491 nm excitation (GFP filter), a return to 561 nm produces a visible signal.
Photoconversion does not appear to be unique to superfolder GFP, as the phenomenon has been reported for other GFPs [e.g., enhanced GFP or GFP(S65T); see http://www.nature.com/articles/srep11771]. We have not yet characterized the conditions for photoconversion, but will report any additional observations in this forum.
Photoconversion shifts GFP into the spectrum which overlaps mCherry and other red FPs, which has the potential to affect double labeling experiments. Our recommendation is that imaging in the red spectrum be performed before GFP imaging, to prevent potential artifacts arising from photoconversion of the latter reporter.
Harold, Kevin, and Andy