Primary Photophysical Processes in Chromoproteins - Fluorescent Proteins: From Understanding to Design - page 58

not involved in the subsequent chemical steps, but that the anionic excited state
formed directly following ESPT is in the appropriate geometry to undergo further
reaction, a geometry which is not directly accessible from the equilibrated anionic
ground state. Most measurements suggest that the quantum yield for photoconver-
sion is low, which is in contrast to the very facile ESPT of avGFP. It seems at least
plausible that the mechanism involves ESPT as a primary step with the resultant
anion either undergoing a second low cross-section step ultimately leading to
photoconversion or a more facile reverse re-protonation in the ground state: such
a mechanism recalls the A

B mechanism introduced for avGFP [ 20 ]. It
seems likely that the role of the ESPT will be resolved through ultrafast and
microsecond transient IR studies, although such measurements on these irreversibly
photoactive proteins present a number of experimental difficulties. Some prelimi-
nary results for kikGR have been presented [ 129 ].
A still more complicated picture is presented by the primary photophysics of the
photochromic CPs, of which Dronpa is the best-characterised example. Dronpa
was engineered by Miyawaki and co-workers from a protein isolated from a coral
[ 130 , 131 ]. It was observed to be very weakly fluorescent following excitation in
the neutral state, but on continued irradiation that state was converted with a
relatively high cross-section to an anionic state which had a high fluorescence
quantum yield. Critically, strong illumination of this emissive anionic state gives
rise to a reverse photoconversion back to the dark state, with a low cross-section.
The protein can be cycled many times through this photochromic cycle, which is
clearly important for the ultraresolution microscopy application [ 131 , 132 ]. Since
the discovery of Dronpa, a number of other photoconvertible proteins have been
reported [ 133 ], including some in which negative switching was achieved, where
excitation of the neutral form switches fluorescence off and irradiation of the
anionic form switches it on [ 24 , 134 ]. Further mutagenesis has been used to control
the rate of the on and off processes, which may vary quite dramatically [ 135 ].
The detailed mechanism of photochromic switching has yet to be fully eluci-
dated. The mechanism is likely to involve a mixture of ESPT, cis - trans isomerisa-
tion, radiationless decay and ground state acid-base equilibria. It is not, however,
clear that there is a common mechanism operating among the different photochro-
mic CPs. In an NMR study, Mizuno et al. showed that the state of protonation of the
chromophore was critical, and in particular that the anchoring of the anionic
(fluorescent) form to the main

! I !

-barrel structure generates a rigid environment for
this state [ 136 ]. In this model, ESPT is an important step [ 137 ], while cis - trans
isomerisation between the off and on states is not critical. However, some X-ray
structure measurements on dark- and light-adapted Dronpa suggested different
isomers of the chromophore in the on and off states, a result that requires some
excited (or ground state) structural reorganisation [ 138 , 139 ]. This observation is
consistent with the reported increased flexibility in the off state, which may well
allow both fast radiationless decay and excited state isomerisation.
The interplay between isomerisation and fluorescence quantum yield is an inter-
esting one. In a series of structural studies of photoswitchable proteins, it was shown
that both planar cis and trans forms of the chromophore could be stabilised, and that

b