Structural and Spectral Response of Green Fluorescent Protein Variants to Changes in pH^,

Abstract

The green fluorescent protein (GFP) from the jellyfish Aequorea victoria has become a useful tool in molecular and cell biology. Recently, it has been found that the fluorescence spectra of most mutants of GFP respond rapidly and reversibly to pH variations, making them useful as probes of intracellular pH. To explore the structural basis for the titration behavior of the popular GFP S65T variant, we determined high-resolution crystal structures at pH 8.0 and 4.6. The structures revealed changes in the hydrogen bond pattern with the chromophore, suggesting that the pH sensitivity derives from protonation of the chromophore phenolate. Mutations were designed in yellow fluorescent protein (S65G/V68L/S72A/T203Y) to change the solvent accessibility (H148G) and to modify polar groups (H148Q, E222Q) near the chromophore. pH titrations of these variants indicate that the chromophore pKa can be modulated over a broad range from 6 to 8, allowing for pH determination from pH 5 to pH 9. Finally, mutagenesis was used to raise the pKa from 6.0 (S65T) to 7.8 (S65T/H148D). Unlike other variants, S65T/H148D exhibits two pH-dependent excitation peaks for green fluorescence with a clean isosbestic point. This raises the interesting possibility of using fluorescence at this isosbestic point as an internal reference. Practical real time in vivo applications in cell and developmental biology are proposed.

An Alternate Proton Acceptor for Excited-State Proton Transfer in Green Fluorescent Protein: Rewiring GFP

“…As observed for S65T, the absorption spectrum of S65T/H148D is sensitive to pH. However, introduction of D148 has caused the pK a observed for S65T (pK a ) 6.0) to increase to 7.8, 21 so that at pH 7.5 a far greater fraction of S65T/H148D is in the neutral form compared to S65T alone. Significantly the introduction of D148 has also caused the λ max of the neutral chromophore to redshift to 411 nm, a shift of 16 nm compared to wtGFP (395 nm) and S65T GFP (394 nm).…”

“…Absorbance and fluorescence data of the GFP mutant, S65T/H148D, have been characterized and are in good agreement with earlier data. 21 The absorption spectrum of the double mutant shows two bands at 411 and 488 nm, which are assigned to the neutral and anionic ground states, respectively ( Figure 2). As observed for S65T, the absorption spectrum of S65T/H148D is sensitive to pH.…”

“…Intriguingly, Remington and colleagues demonstrated that replacement of H148 with an aspartate results in the recovery of green fluorescence following excitation of the A state. 21 H148 is situated close to the chromophore's phenolic hydroxyl residue, leading to the suggestion that the carboxylate of D148 is the proton acceptor for ESPT in this mutant. In the current work we have explored the excited-state dynamics of S65T/H148D GFP using steady-state and ultrafast time-resolved fluorescence (TRF) and ultrafast TR infrared (TRIR) spectroscopies, and we have contrasted them with the single mutant S65T GFP.…”

“…Unlike wtGFP, the chromophore in S65T is sensitive to pH, and conversion between A and B forms of the chromophore occurs with a pK a of 6. 21,22 Thus, S65T GFP acts as a pH sensor. In addition, at physiological pH the anionic B form predominates, thereby increasing the fluorescent signal resulting from direct excitation of this form of the chromophore.…”

“…The mutation of S65 to a threonine alters contacts between the side chain of this residue and E222, causing the side chain of the latter to move so that it can no longer participate in ESPT. 18,21 Thus, for S65T, the I* state cannot form, and excitation into A results in only weak fluorescence directly from A*.…”