Expression of mTurquoise and its variants in E. coli was measured at 37 °C using a fluorescence microplate reader, as described9 (link), except that we used the pRSET plasmids for expression without induction. For brightness studies in mammalian cells, mTurquoise variants were excised from the pRSET vector using NheI and BsrGI and ligated into the CFP-2A-SYFP2 co-expression vector that was partially digested with NheI and BsrGI, and brightness analysis was performed as described10 (link) 2 days after transfection. For expression of mTurquoise variants in HeLa cells, we replaced the fluorescent protein from pSCFP3A-C1 with the mTurquoise variant using AgeI/BsrGI restriction sites. Photostability was measured in living HeLa cells by continuous illumination with light from a 100 W mercury lamp that was passed through a 436/20-nm excitation filter and reflected onto the sample with a 455DCLP dichroic mirror (Chroma). A ×63 oil immersion objective (Plan Apochromat NA 1.4) was used and a light intensity of 1.4–1.6 W cm−2 was determined when a ND1.3 grey filter was used, whereas a light intensity of 22–23 W cm−2 was determined without grey filter. To evaluate photostability under laser scanning conditions, we used HeLa cells expressing CFP variants tagged with histone 2A. A Nikon A1 laser scanning microscope was used and cells were imaged using a Nikon Plan Apo VC ×60 oil objective and a zoom factor of 6. Excitation was at 443 nm, and fluorescence was passed through a completely opened pinhole and a 482/35-nm bandpass filter. All parameters, including detector gain, laser intensity, and scan speed were identical between samples. Spectral imaging microscopy was performed on single cells as described10 (link).