Novel dual emission, pH-sensitive variants of the green fluorescent protein (GFP) have been constructed and are suitable for ratiometric emission measurements in vivo. This new class of GFPs, termed deGPFs, results from substitution of wild-type residue 65 with threonine and residues 148 and/or 203 with cysteine. deGFPs display pK(a) values ranging from 6.8 to 8.0 and emission that switches from a green form (lambda(max) approximately 515 nm) to a blue form (lambda(max) approximately 460 nm) with acidifying pH. In this report we analyze in most detail the deGFP1 variant (S65T/H148G/T203C, pK(a) approximately 8.0) and the deGFP4 variant (S65T/C48S/H148C/T203C, pK(a) approximately 7.3). In the following paper [McAnaney, T. B., Park, E. S., Hanson, G. T., Remington, S. J., and Boxer, S. G. (2002) Biochemistry 41, 15489-15494], data obtained by ultrafast fluorescence upconversion spectroscopy can be described by a kinetic model that includes an excited-state proton-transfer pathway at high pH but not at low pH. Crystal structure analyses of deGFP1 at high-pH and low-pH conformations were performed to elucidate the basis for the dual emission characteristics. At low pH the structure does not contain a hydrogen bond network that would support rapid transfer of a proton from the excited state of the neutral chromophore to a suitable acceptor; hence blue emission is observed. At high pH, backbone rearrangements induced by changes in the associated hydrogen bond network permit excited-state proton transfer from the excited state of the neutral chromophore to the bulk solvent via Ser147 and bound water molecules, resulting in green emission from the anionic chromophore. Comparative analysis suggests that the basis for dual emission is elimination of the wild-type proton-transfer network by the S65T substitution, a general reduction in hydrogen-bonding opportunities, and a concomitant increase in the hydrophobic nature of the chromophore environment resulting from the cysteine substitutions. We evaluated the suitability of the deGFP4 variant for intracellular pH measurements in mammalian cells by transient expression in PS120 fibroblasts. The responses of deGFP4 and a commercially available pH-sensitive dye, SNARF-1, to changes in pH were compared in the same cells. Results show that the dynamic range of the emission ratio change is comparable between the two pH sensors over the range examined. Two-photon excitation was found to elicit a better deGFP4 fluorescent signal above cellular autofluorescence when compared to conventional confocal microscopy. Given their favorable optical characteristics, suitable pK(a)'s for the physiological pH range, and suitability for ratiometric measurements, dual emission GFPs should make excellent probes for studying pH in vivo.

The betaine transporter (BGT1) protects cells in the hypertonic renal inner medulla by mediating uptake and accumulation of the osmolyte betaine. Transcriptional regulation plays an essential role in upregulation of BGT1 transport when renal cells are exposed to hypertonic medium for 24 h. Posttranscriptional regulation of the BGT1 protein is largely unexplored. We have investigated the distribution of BGT1 protein in live cells after transfection with BGT1 tagged with enhanced green fluorescent protein (EGFP). Fusion of EGFP to the NH2 terminus of BGT1 produced a fusion protein (EGFP-BGT) with transport properties identical to normal BGT1, as determined by ion dependence, inhibitor sensitivity, and apparent Km for GABA. Confocal microscopy of EGFP-BGT fluorescence in transfected Madin-Darby canine kidney (MDCK) cells showed that hypertonic stress for 24 h induced a shift in subcellular distribution from cytoplasm to plasma membrane. This was confirmed by colocalization with anti-BGT1 antibody staining. In fibroblasts, transfected EGFP-BGT caused increased transport in response to hypertonic stress. The activation of transport was not accompanied by increased expression of EGFP-BGT, as determined by Western blotting. Membrane insertion of EGFP-BGT protein in MDCK cells began within 2-3 h after onset of hypertonic stress and was blocked by cycloheximide. We conclude that posttranscriptional regulation of BGT1 is essential for adaptation to hypertonic stress and that insertion of BGT1 protein to the plasma membrane may require accessory proteins.