Proton Travel in Green Fluorescent Protein - Fluorescent Proteins: From Understanding to Design - page 167

experiments in H 2 O and in D 2 O. If the rate constants of photo-physical processes
are slowed down considerably, this is commonly taken as clear evidence for the
involvement of protons and is characterized by the kinetic isotope effect (KIE

¼
k D / k H ). In the case of green fluorescence protein (GFP), this is exactly what was
done in the classical paper by Chattoraj et al., which demonstrated the existence of
ESPT upon photo-excitation [ 9 ]. There, the decay of the photo-excited species A*
and the rise of the I* are slowed down by a factor of about 5 when performing the
experiments in D 2 O relative to H 2 O.
Spatial mapping of proton-transfer pathways in proteins is the realm of Fourier-
transformed infrared spectroscopy (FTIR) and has been applied to map out PT
pathways in bacteriorhodopsin, the bacterial photosynthetic reaction center, and in
ras p21 [ 10 ]. For this, one typically takes time-resolved difference spectra between
wild-type and mutant protein. GFP is an ideal system for this technique because the
chromophore is naturally built into the proton barrel so that one can simultaneously
apply optical and infrared spectroscopy. So far, FTIR analysis of GFP has focused
on mapping the immediate chromophore surrounding [ 11 ].

2 Fluorescence Autocorrelation of GFP

Fluorescence correlation spectroscopy (FCS) experiments have been used to probe
the fluorescence fluctuations of GFP proteins over time [ 12 ]. For example, one may
probe whether a GFP protein that predominantly contains a neutral chromophore at
time t 0 still contains a neutral chromophore after time t 0 + t. When fluorescence can
arise only from the neutral chromophore state, then the reversible loss of a proton
leads to a reversible loss of fluorescence thus resulting in fluorescence fluctuations.
The fluctuations are subsequently analyzed by its autocorrelation G ( t ) . Changes of
the protonation state of the chromophore may occur either due to proton exchange
with other titratable groups of the protein or via proton exchange with bulk solvent
[ 13 ]. Interestingly, the decay of the autocorrelation of the EGFP chromophore on
the sub-millisecond time scale showed a clear pH dependence [ 12 ] (Fig. 1 ). This pH
dependence of the decay time is a clear evidence that the protein interior is in
dynamic exchange with the surrounding proton reservoir on a micro- to millisecond
time scale. This on-off fluorescence flickering of GFP mutants is influenced by
several environmental factors. The pH effect, attributed to reversible external
protonation of the chromophore, was discussed for EGFP [ 12 ]. In addition, a
fraction ( 13%) of EGFP molecules showed a pH-insensitive flicker that seemed

dependent on illumination intensity. Schwille and coworkers presented thorough
intensity- and pH-dependent FCS studies of flicker dynamics in the yellow-shifted
mutants T203Y and T203F [ 14 ] that allowed an effective separation of these two
effects, pH vs. intensity. So far, it is unclear whether this pH dependence is of
functional relevance. Yet, the coupling of internal proton transfer to the chromo-
phore fluorescence makes GFP a unique probe to characterize and understand such
proton-transfer processes.