Proton Travel in Green Fluorescent Protein

Green fluorescence protein (GFP) wild type and some of its mutants undergo excited state proton transfer between the chromophore and the nearby Glu222 residue. This process has been covered in detail in the chapter written by Stephen Meech. Apart from this ultrafast photochemical reaction, multiple other proton-transfer processes take place in the GFP protein matrix, and these will be covered in this chapter. For example, proton exchange between the chromophore and the nearby bulk solvent may occur via His148 that is located in hydrogen-bonding distance from the chromophore and provides direct access to the bulk solvent. Moreover, two extended proton-transfer wires including titratable residues as well as a number of buried water molecules connect the chromophore to the protein surface. Based on a recent high-resolution X-ray structure of GFP, all titratable groups of the protein could be placed in one of these two large hydrogen-bonding clusters, suggesting that a multitude of proton-transfer processes can occur in the GFP matrix at any moment in time. While it is quite likely that similar proton pathways also exist in other soluble and membrane proteins, they are much harder to study. GFP is an exciting model system for monitoring those processes as they often directly affect the chromophore photophysics. The dynamics of proton exchange inside the GFP barrel and with bulk solvent has thus been characterized by fluorescence correlation spectroscopy (FCS) of the chromophore fluorescence and by pH-jump experiments. These studies showed that the autocorrelation of the chromophore fluorescence is affected either by pH-independent processes on microsecond to millisecond time scales or by pH-dependent processes on similar time scales. The former ones are likely proton equilibria occurring within the GFP barrel, and the latter ones are likely exchange processes with the solvent. Biomolecular simulation methods are now being developed, which will soon allow accessing such time scales by computational means. Then, we will hopefully be able to connect the spectroscopic findings with dynamic atomistic simulations of proton-transfer dynamics.