Abstract
Fluorescent protein (FP) tagging is widely used in imaging experiments to investigate the subcellular distribution of proteins. However, because the fluorescence of most FP chromophores is quenched upon their protonation, their fluorescence intensities are dependent on their pKas and on the environmental pH. Thus, the concentration of a protein tagged with EGFP (pKa = 6.0) is dramatically underestimated in the lysosomal lumen (pH ~4.7) compared to that of the same protein tagged with mCherry (pKa = 4.5). In this study, we examined the effect of differential FP tagging on the apparent subcellular distribution of several proteins that reside on the cytoplasmic surfaces of secretory/endocytic organelles. Due to the presumed uniformity of cytoplasmic conditions (pH ~7.2-7.4), we expected to find essentially complete overlap of fluorescent signals, regardless of the nature of the fused FP. However, we were surprised to observe significant discrepancies in the apparent distributions of a subset of proteins tagged with EGFP vs. mCherry (Pearson's correlation coefficients of about 0.80). These discrepancies were not evident when comparing proteins tagged with mCherry vs. other FPs with low pKas (e.g., mTurquoise (pKa = 4.5), mCerulean (pKa = 3.2)) (Pearson's correlation coefficients of about 0.90-0.95). Our results suggest that FP tags may be sensitive to the microenvironments on the cytoplasmic surfaces of different organelles.