Ultrafast dynamics of isolated model photoactive yellow protein chromophores: "Chemical perturbation theory" in the laboratory

Article abstract of DOI:10.1021/jp045763d

Pump-probe and pump-dump probe experiments have been performed on several isolated model chromophores of the photoactive yellow protein (PYP). The observed transient absorption spectra are discussed in terms of the spectral signatures ascribed to solvation, excited-state twisting, and vibrational relaxation. It is observed that the protonation state has a profound effect on the excited-state lifetime of p-coumaric acid. Pigments with ester groups on the coumaryl tail end and charged phenolic moieties show dynamics that are significantly different from those of other pigments. Here, an unrelaxed ground-state intermediate could be observed in pump-probe signals. A similar intermediate could be identified in the sinapinic acid and in isomerization-locked chromophores by means of pump-dump probe spectroscopy; however, in these compounds it is less pronounced and could be due to ground-state solvation and/or vibrational relaxation. Because of strong protonation-state dependencies and the effect of electron donor groups, it is argued that charge redistribution upon excitation determines the twisting reaction pathway, possibly through interaction with the environment. It is suggested that the same pathway may be responsible for the initiation of the photocycle in native PYP.

Full text of DOI:10.1021/jp045763d

J. Phys. Chem. B 2005, 109, 4197-4208
4197
Ultrafast Dynamics of Isolated Model Photoactive Yellow Protein Chromophores:
“
Chemical Perturbation Theory” in the Laboratory
,
†
,†,§
‡
†
Mikas Vengris,* Delmar S. Larsen,* Michael A. van der Horst, Olaf F. A. Larsen,
‡
†
Klaas J. Hellingwerf, and Rienk van Grondelle
Faculty of Sciences, Vrije UniVersiteit Amsterdam, De Boelelaan 1081, 1081 HV Amsterdam, The Netherlands,
and Department of Microbiology, Swammerdam Institute for Life Sciences, UniVersity of Amsterdam,
Nieuwe Achtergracht 166, 1018 WS Amsterdam, The Netherlands
ReceiVed: September 17, 2004; In Final Form: December 6, 2004
Pump-probe and pump-dump probe experiments have been performed on several isolated model
chromophores of the photoactive yellow protein (PYP). The observed transient absorption spectra are discussed
in terms of the spectral signatures ascribed to solvation, excited-state twisting, and vibrational relaxation. It
is observed that the protonation state has a profound effect on the excited-state lifetime of p-coumaric acid.
Pigments with ester groups on the coumaryl tail end and charged phenolic moieties show dynamics that are
significantly different from those of other pigments. Here, an unrelaxed ground-state intermediate could be
observed in pump-probe signals. A similar intermediate could be identified in the sinapinic acid and in
isomerization-locked chromophores by means of pump-dump probe spectroscopy; however, in these
compounds it is less pronounced and could be due to ground-state solvation and/or vibrational relaxation.
Because of strong protonation-state dependencies and the effect of electron donor groups, it is argued that
charge redistribution upon excitation determines the twisting reaction pathway, possibly through interaction
with the environment. It is suggested that the same pathway may be responsible for the initiation of the
photocycle in native PYP.
1. Introduction
ultimately define the driving force for the reaction. The solvent
environment also plays a key part in determining the course of
the reaction. In biological systems, an even more important role
Light-induced reactions, especially those that include struc-
tural changes in molecules, have long been of interest for
scientists. Understanding these reactions represents not only a
scientific challenge but also has significant relevance for industry
and biology. The molecules that undergo photoinduced reactions
have potential applications in such areas as optical memory,
optoelectronics, solar cells, molecular switches, etc.1-3 Studying
these molecules relates to a multitude of condensed-phase
processes of fundamental importance, including charge redis-
tribution, solvation, trans-cis isomerization, and inter- and
intramolecular vibrational relaxation. Important processes of life,
is played by the protein environment because the protein reacts
to the structural change of the pigment and performs its relevant
biological function (e.g., in bacteriorhodopsin, a proton is
transferred across the membrane, and photoactive yellow protein
(PYP) changes its conformation to alter its interaction(s) with
the signaling partner).
One of the better characterized proteins with a biological
function based on an isomerization photoreaction of an intrinsi-
cally bound chromophore, is the PYP. Because of its small size
and excellent stability, it has become a model system for
4
5
including human vision, bacterial photosynthesis, plant pho-
7
6
7
studying biological light-induced signal transduction. PYP is
totropism, and bacterial phototaxis are also initiated by these
reactions.
a small (14 kDa) water-soluble protein belonging to the PAS
domain family and was originally discovered in the photosyn-
To be successful, the rates of photoinduced excited state
reactions must compete effectively with the lifetime of electronic
excitation in these molecules (usually of the order of several
ns for singlet excited states), which puts these reactions in the
realm of ultrafast laser spectroscopy. Numerous ultrafast studies
1
1
thetic bacterium Halorhodospira halophila, where it initiates
a negative phototactic response to blue-light illumination.12
A
significant amount of research on this protein was triggered by
the resolution of its stationary (both room temperature and
1
3
14-16
8
cryotrapped) and time-resolved X-ray structures.
These
on “classical” isomerizing molecules such as trans-stilbene,
9
10
structures confirmed the earlier assertion that PYP contained a
auramine, and the protonated Schiff base containing polyenes
have revealed that the rates of these reactions are in the (sub)-
picosecond time domain. The answer to the question as to how
the reactions are actually proceeding is encrypted in the excited-
state potential-energy surfaces (PESs) of these molecules, which
1
7
p-coumaric acid chromophore (pCA) that is covalently
bound to the protein backbone via a thioester linkage to the
cysteine-69 residue. This pigment is buried inside a protein
pocket that protects it from the solvent environment; in the dark-
adapted ground state, its phenolic moiety is deprotonated, and
the negative charge is stabilized by the network of hydrogen
bonds with neighboring amino acid residues. Upon blue-light
excitation of the pCA pigment, a complex photocycle is initiated
that exhibits a dynamic sequence of intermediates occurring on
multiple time scales from tens of femtoseconds to seconds,7
*
Authors to whom correspondence may be addressed. E-mail: mikasv@
nat.vu.nl (M.V.); dlarsen@usc.edu (D.S.L.).
†
Vrije Universiteit Amsterdam.
University of Amsterdam.
Current address: Department of Chemistry, University of Southern
‡
§
California, Los Angeles, CA 90089.
1
0.1021/jp045763d CCC: $30.25 © 2005 American Chemical Society
Published on Web 02/15/2005

4198 J. Phys. Chem. B, Vol. 109, No. 9, 2005
Vengris et al.
Figure 1. Absorption spectra and structural formulas of the PYP model chromophores investigated in this study. In all panels, except panel F, the
solid, dashed, and dotted lines represent the spectrum measured at Ph ) 10.5, 7, and 2, respectively. (A) pCA; (B) TMpCA; (C) OMpCA; (D)
FerFA; (E) SinFA. (F) FAL is the solid line, while thiomethyl locked TML is demarked by a dashed line.
including a trans-cis isomerization and protonation of the pCA
pigments (Figure 1) in aqueous solution are contrasted and
compared. While all the studied chromophores share the same
carbon backbone, which is identical to the native PYP chro-
1
8,19
chromophore and a partial unfolding of the protein.
The
partially unfolded form of PYP is the presumed signaling state
responsible for biochemical signal transduction.20 The original
configuration of the protein and the chromophore is recovered
on a time scale of 0.1-1 s.
38
mophore, they differ in the nature of their substituents and
protonation state. The thiomethyl ester derivatives of pCA
(TMpCA) explore the role of the thioester linkage, which is
also present in the PYP protein (provided by the Cys69 residue),
on the ensuing dynamics, and these observations are contrasted
with the oxy-methyl-pCA (OMpCA) chromophores, where this
ester linkage contains the lighter oxygen atom instead. Ferulic
(FerFA) and sinapinic (SinFA) acids contain one and two
methoxy groups, respectively, attached to the phenolic ring of
the pigment, which may sterically hinder the twisting reaction
and/or may function as electron-donating groups which alter
the charge distribution in the conjugated system and affect the
charge-transfer character of the electronic transition. The data
collected for these free-twisting chromophores are contrasted
with locked chromophores, where the possibility of twisting/
isomerization is prevented by the introduction of additional
covalent bonds. The free-acid-locked (FAL) and the thiomethyl-
locked (TML) derivatives are useful to identify observed
dynamical features that originate from processes other than
twisting, such as solvation and vibrational relaxation.
Recently, a number of spectroscopic ultrafast studies have
2
1-25
focused on investigating the initial dynamics of PYP.
Furthermore, mutants with genetically engineered changes in
the chromophore-binding pocket, along with hybrid proteins
21-23,25-30
containing non-native pigments, have been examined.
Changing the amino acid residues in the protein pocket
consistently appears to slow the excited-state decay,2 whereas
the incorporation of a nonisomerizing pigment analogue into
the protein has been shown to completely block the signaling-
7
7,31
state formation. Some mutations, and especially a combina-
tion of a mutation plus a replacement of the native pigment,
32
resulted in a dramatic change in the color of the protein. Some
of the key questions addressed in these studies was to what
extent the photoinduced dynamics in PYP is dictated by the
dynamics inherent to the chromophore and how this dynamics
is tuned by the protein environment, or by the interactions
between the two. Previously, several potential model PYP
chromophores have been investigated in solution with ultrafast
Because the chromophore binding pocket in PYP is tuned to
generate a high pH (∼9.8), experiments on the pigments were
performed at different pH values to investigate the effect of
protonation on the light-induced dynamics. Thus, by collecting
these data on a variety of chromophores and under various
conditions, a sort of “chemical perturbation theory” has been
applied in the laboratory to explore the intrinsic dynamics of
the PYP chromophore.
spectroscopic techniques.3
3-36
In model chromophores contain-
ing a thioester moiety, an unrelaxed ground-state intermediate
has been observed that was ascribed to a twisted conformation
of the chromophore, which subsequently returned back into the
equilibrated ground-state structure. Similar spectroscopic fea-
3
7
tures have also been observed in the native PYP system,
suggesting that the study of specific isolated chromophores can
21,25,34,37
be useful for exploring the chromophore dynamics in PYP.
In this paper, we present a systematic study of the dynamics
of several PYP model compounds, with an emphasis on
identifying the dynamical effects arising from changing the
structure, pH, or excitation wavelength. Dispersed pump-probe
2
. Experimental Section
2.1. Samples. The pCA, SinFA, and FerFA chromophores
were purchased from Sigma-Aldrich, while FAL was purchased
from Molecular Probes; these pigments were used without
(PP) and pump-dump probe (PDP) data on the seven different

Photoactive Yellow Protein Chromophores
J. Phys. Chem. B, Vol. 109, No. 9, 2005 4199
further purification. The thiomethyl derivatives were synthesized
from their respective free acids (via the 1,1-carbonyldiimidazole
intermediate) and sodium thiomethoxide (Acros Organics, Geel,
Belgium). Equimolar amounts were mixed and allowed to react
overnight at room temperature. The chromophores were purified
with a silica gel column by washing with two column volumes
of petroleum ether and then eluted with a 1:1 (v/v) ethyl acetate/
petroleum ether mixture. For the synthesis of OMpCA (methyl-
coumarate or 4-hydroxycinnamic methylester), 1 mmol of
not necessarily represent the real spectra of the excited-, ground-,
and product-state species that occur in the experiment.4
1,42
As
opposed to the ubiquitous application of fitting measured data
to a “sum of exponential decays,” which essentially is a parallel
model, the estimated SADS resulting from a sequential model
resemble the real PP spectra measured at certain delay times.
Thus, sequential SADS are often more intuitive to interpret than
the raw data directly. For a more detailed discussion about the
applicability of various models and the interpretation of the
results of global and target analysis the reader is referred to
4
-hydroxy-cinnamic acid was dissolved in 10 mL of methanol,
and a catalytic amount of concentrated sulfuric acid was added
a few drops). The reaction mixture was stirred overnight at
4
1-43
references.
(
In all the fits presented here, a component shorter than the
time resolution of the instrument (lifetime ca. 100 fs for 395
nm excitation and ca. 250 fs for 310-350 nm excitation) was
introduced to adequately describe the data. The SADS corre-
sponding to this component, however, is significantly contami-
nated by the cross-phase modulation artifact between the pump
room temperature and evaporated under reduced pressure, and
the product was further purified by recrystallization from a
petroleum ether-ethyl acetate mixture.
The identity and purity of the synthesized compounds was
confirmed with NMR and mass spectrometry. The samples were
buffered with 50 mM CAPS buffer at a pH of 10.5 or 50 mM
phosphate buffer for pH ) 7 and 2 and were prepared with
optical densities (OD) between 0.5 and 0.7 per mm at the
absorption maximum (Figure 1).
4
3,44
and probe pulses
and has been omitted from the presented
global analysis figures.
3
. Results
The experimental results section is organized in the following
2
.2. Ultrafast Setup. The PP setup has been described in
36,39
manner: first, the raw PP data of the investigated pigments are
discussed. Because of space considerations, we only discuss the
data explicitly for two systems: pCA at pH 10.5 and SinFA at
pH 10.5, and for the other pigments, only the raw kinetic traces
measured at the maximum of the stimulated emission (SE) band
are presented. Additionally to the raw data, we present the
corresponding SADS resulting from the sequential global
analysis fit to the data, which illustrates the spectral aspects of
the ensuing dynamics. Next, the results of wavelength-dependent
PP experiments and finally the PDP data on SinFA and locked
chromophores are presented.
detail earlier
and has been modified to generate excitation
light in the near-UV region. The basis of the system is a 1-kHz
amplified Ti:sapphire system (BMI R) delivering 400-µJ, 60-
fs, and 790-nm pulses. The dispersed PP experiments presented
were performed at two excitation wavelengths: 395 nm and in
the near-UV region (between 310 and 350 nm). The 395-nm
pump pulses were generated by frequency doubling a portion
of the amplified 790-nm light in a 0.5 mm thick BBO crystal,
while the UV pump pulses were generated by mixing the 475-
6
25 nm output of a home-built 395-nm-pumped noncollinear
optical parametric amplifier (NOPA) with the fundamental in
the 0.4 mm thick BBO crystal. For three-pulse PDP experiments,
3.1. Transient Absorption. The pCA and SinFA PP datasets
measured at pH 10.5 are compared in Figure 2. The excitation
wavelength was 330 nm for pCA and 395 nm for SinFA. Every
point in the kinetic traces (parts A and C of Figure 2) is taken
from a measured time-gated spectrum, several of which at
different probe delay times are shown in parts B and D of Figure
the 505-nm beam from the NOPA was introduced as a dump
pulse.37 The white-light continuum, used as the broad-band
probe pulse, was produced by focusing a weak 790-nm beam
into a slowly translating CaF2 crystal. The delay between pump
and probe pulses was varied using motorized, computer-
controlled translation stages (Newport). The samples were
pumped through a rapidly translating 1-mm quartz flow cell.
The polarization of the pump (and dump) pulse was kept at
magic angle (54.7°) with respect to the probe pulse. The group
velocity dispersion (GVD) of the white light was estimated from
the measurement of the cross-phase modulation artifact in water
and was further refined in the global fitting procedure, and the
data were corrected for GVD prior to presentation.
2
. The spectral features of both compounds are qualitatively
similar: a positive induced absorption (IA) band is observed at
high probe energies (ca. 360 nm for pCA and 400 nm for
SinFA), which overlaps the ground-state absorption, while a
negative SE band peaks at lower probe energies (430 nm for
pCA and 485 nm for SinFA). For both chromophores, the IA
obscures the expected ground-state bleach (parts A and E of
Figure 1). The time-gated spectra in parts B and D of Figure 2
exhibit a dynamic red-shifting of the SE on the <2-ps time scale.
From the comparison of the spectra obtained at 500 fs and 2 ps
for pCA (Figure 2A), a clear decay of the IA band is observed
that is not correlated with a loss in the SE band. However, on
the longer time scales (> 2 ps), the decay of the SE and IA
bands is identical, indicating that this IA is due to the excited-
state absorption (ESA). In both samples, the kinetic traces
around the zero crossing point of the spectrum (the spectral
region, where ∆OD ) 0 at different delay times, see the spectra
in parts B and D of Figure 2) show a behavior different from
either the SE or ESA bands; initially the signal is negative, but
after 1 ps, it crosses zero, reaches the positive maximum at
around 3 ps, and then decays to zero. At these wavelengths
that correspond to the zero-crossing point of the spectrum, the
kinetic traces are very sensitive to spectral shifts and the
appearance of new bands; the observed behavior is indicative
of a shifting SE band or a growth and decay of a product
absorption.
The collected data constitute a three-dimensional surface with
ca. 250 wavelength points and ca. 100 time points. In the case
of PDP experiments, it actually consists of three surfaces,
corresponding to PP (pump pulse on, dump pulse off), DP (pump
pulse off, dump pulse on), and PDP (both pump and dump
pulses are on) data. The wavelength resolution of the data is
ca. 1 nm, and time resolution is ∼130 fs with 395-nm excitation
and ∼370 fs with UV excitation respectively, both with an
average noise level of ∼1 mOD.
2
.3. Data Analysis. The PP data was fitted globally using
τ1
τ2
τ3
40,41
an evolutionary model (i.e., A f B f C f ...),
which
reduces the often unwieldy time-resolved and wavelength-
resolved data to a set of a small number of spectra (typically
around five), which evolve into each other with consecutively
decreasing lifetimes of τ1, τ2, etc. Time evolution of the spectrum
can thus be described in a concise parametric manner; however,
the obtained species-associated difference spectra (SADS) do

4200 J. Phys. Chem. B, Vol. 109, No. 9, 2005
Vengris et al.
Figure 2. Dispersion-corrected PP data on pCA (A and B) and SinFA (C and D) at pH ) 10.5. (A and C) Kinetic traces at the ESA band (squares),
around the zero-crossing point (circles), and SE band (triangles). The traces at the zero crossing point have been multiplied 3-fold, and the traces
at ESA and SE bands have been offset by 25 and -25 mOD (10 and -10 mOD for SinFA), respectively. The probe wavelengths for the traces are
shown in the legends. The solid lines show the global analysis fit through the data. Note the break in the time axis. (B and D) Time-gated spectra
measured at the probe delay times shown in the legends. In SinFA data (D), the spectral region around 395 nm was corrupted by the scattering of
the pump light and is not shown.
Figure 3. (A) Kinetic traces measured at the SE band of pCA at pH ) 2 (open squares), 7 (filled circles), and 10.5 (open triangles). The excitation
wavelength was 310 nm, and the probe wavelength for the shown traces was 430 nm. Solid lines show the global analysis fit to the data. (B, C,
and D) SADS resulting from the sequential model fit to the data measured on pCA at pH ) 2, 7, and 10.5, respectively. The estimated time
constants for the spectral evolution are shown in the legends.
The pH dependence of the spectra and kinetics for pCA is
presented in Figure 3. The decay of the SE band at pH ) 2 and
is nearly pH independent (slightly faster at pH ) 2). However,
at pH ) 10.5, a significant decrease in the decay rate is observed
Figure 3A, open triangles). At pH ) 2 and 7, essentially all
the transient absorption signal decays on a 2.6-ps time scale,
whereas at pH ) 10.5 the decay takes 9.4 ps (parts B, C, and
D of Figure 3). For all three pHs, the SADS exhibit a faster
initial decay in the IA band than in SE band as observed in
Figure 2B. A slight red-shift of the SE is observed with a time
constant of ca. 500 fs at pH ) 2 and 10.5; however, it is nearly
absent at pH ) 7. At all pH values, a nondecaying component
is also observed; it has a near-zero amplitude and exhibits only
an induced absorption.
OMpCA and TMpCA exhibit much richer dynamics than
pCA. The dispersed PP data on these compounds at pH ) 7
(excitation wavelength 310 nm) and 10.5 (excitation wavelength
395 nm) are shown in Figure 4. The OMpCA data at pH ) 7
(Figure 4C) closely resemble those obtained with pCA (Figure
3): a 12-nm red-shift of SE is observed on a 570-fs time scale
along with a partial decay of the ESA band, and essentially all
the SE decays with a 2-ps lifetime (although a very weak 8-ps
component is also observed). Similar behavior is observed in
TMpCA at pH ) 7 but with a slightly longer 3.7-ps lifetime.
In contrast to the neutral pH conditions, these chromophores
show distinctly different dynamics at pH ) 10.5. For TMpCA,
the dynamics has been investigated in detail and presented
7
(
3
5,36
elsewhere,
and the same data is reproduced here for

Photoactive Yellow Protein Chromophores
J. Phys. Chem. B, Vol. 109, No. 9, 2005 4201
Figure 4. Time-resolved kinetic traces measured at the SE band of TMpCA (A) and OmpCA (B) at pH ) 7 (squares) and 10.5 (circles) along with
the global analysis fit to the data (solid lines). Excitation wavelength: 310 nm for pH ) 7 and 395 nm for pH ) 10. Probe wavelengths are shown
in the legends. (C and E) SADS resulting from the sequential model fit to the data measured on OMpCA at pH ) 7 and 10.5, respectively. (D and
F) TMpCA SADS at pH ) 7 and 10.5, respectively. The estimated time constants for the spectral evolution are shown in the legends.
comparison. Both pigments exhibit sharper and more asym-
metric SE bands than pCA at all pH values and than their own
counterparts at pH ) 7. In both OMpCA and TMpCA, a
significant red-shifting of the SE bands is observed at the time
scale from 100 to 860 fs. In OMpCA, the resolution of the fast
time scales is poorer than in TMpCA because of the smaller
spectral overlap between the 395-nm excitation pulse and the
ground-state absorption (parts C and D of Figure 1), which
results in a greater contamination of the data with the cross-
phase modulation artifact. The decay of SE is not significantly
influenced by the pH (parts A and B of Figure 4). However,
the most striking observation is that, along with the red-shifting
and decay of SE, a rise of an IA band is observed peaking at
ca. 390 nm for OMpCA (Figure 4E) and 440 nm for TMpCA
19 ps in FerFA and 26 ps in SinFA. At pH ) 10.5, two decay
components are resolved in FerFA (8.2 and 44 ps); SinFA decay
is largely monoexponential with an 18-ps lifetime. The SADS
representing SE decay (44 ps for FerFA and 18 ps for SinFA)
exhibit the most red-shifted SE band (parts E and F of
Figure 5) and feature a shoulder on the red wing of the ESA
band (ca. 425 nm). This spectral region was further investigated
by PDP experiments (see below).
The FAL and TML chromophores, which have the C1sCâ
single bond and the CRdCâ double bond of their coumaryl
tail locked, are essentially fluorescent dyes with a lifetime of
the order of several nanoseconds (Figure 6). The SADS of both
chromophores (parts B and C of Figure 6) show a dynamic red-
shift of the SE with characteristic times of 320 fs and 2.2 ps in
FAL and 200 fs and 1.27 ps in TML (the process with a lifetime
shorter than our time resolution was also observed but is
obscured by the cross-phase modulation artifact and not included
in the graphs). In both compounds, clear monoexponential
kinetic decays are observed, with time constants of 3.8 ns for
FAL and 2.1 ns for TML.
(Figure 4F). The kinetic traces taken at the maximum of this
IA band (not shown) exhibit similar features to those observed
in the zero-crossing traces of pCA and SinFA at pH ) 10.5
(parts A and C of Figure 2).
The pigments with methoxy group(s) attached to the phenolic
ring (FerFA and SinFA) exhibit significantly longer lifetimes
than pCA, TmpCA, and OMpCA; the traces at the peak of SE
band of these pigments at different pH values are contrasted in
Figure 5. Even though the excited-state decay is distinctly longer
In most of the chromophores, a weak component exhibiting
no temporal dynamics in the time window of the experiment
(4 ns) is observed. This “terminal spectrum”, which is most
noticeable in the SinFA dataset at pH ) 10.5 (Figure 5F),
exhibits a broad induced absorption band peaking to the red of
our probe window (>650 nm) that is often accompanied by a
narrower IA peak around or below 400 nm (parts C-F of
Figure 5). These bands have been observed in TMpCA and other
(
∼20 ps), the spectral features observed in the SADS are similar
to those observed for the other pigments. A clear red-shift of
the SE is observed in these chromophores that takes place from
6
00 to 2.8 ps, depending on the sample. Similar to pCA, in
FerFA and SinFA at pH ) 10.5 a decay of the IA band is
observed along with a shift; this decay, however, is absent in
the pH ) 7 data, where the IA band remains unchanged during
the first several ps (parts C and D of Figure 5). At pH ) 7, the
SE decay of both pigments is largely monoexponential and takes
3
6,45
chromophores before
and are ascribed to the photoproducts
arising from a two-photon ionization of the chromophore:
solvated electrons (broad band peaking at 720 nm) and
corresponding radicals (narrow induced absorption band).

4202 J. Phys. Chem. B, Vol. 109, No. 9, 2005
Vengris et al.
Figure 5. Time-resolved kinetic traces measured at the SE band of FerFA (A) and SinFA (B) at pH ) 7 (filled squares) and pH ) 10.5 (open
circles) along with the global analysis fit to the data (solid lines). Excitation wavelength: 318 nm for the data measured at pH ) 7 and 395 nm for
pH ) 10.5. Probe wavelengths are shown in the legends. (C and E) SADS resulting from the sequential model fit to the data measured on FerFA
at pH ) 7 and 10.5, respectively. (D and F) SinFA SADS at pH ) 7 and 10.5, respectively. The estimated time constants for the spectral evolution
are shown in the legends.
3
.2. Excitation Wavelength Dependence. The observed
dynamics in the pigments may be influenced by the amount of
excess energy introduced into the molecule (i.e., they may
depend on where in the absorption band the molecule is excited).
For many of these chromophores, it is often impossible to select
optimal wavelengths for excitation. The excess vibronic energy
administered to the molecule is redistributed among the mo-
lecular vibrational degrees of freedom and subsequently trans-
ferred to the environment, and these processes may be reflected
in the electronic transient spectra. In some cases, the absorption
band of the pigment may be due to several closely spaced
excited electronic states,46 and the relaxation between those
would depend on the initial populations created by the excitation
pulse (in other words, the observed dynamics would also be
sensitive to the excitation wavelength). To address this issue,
dual-excitation PP experiments47 were performed on FerFA and
SinFA at pH ) 10.5, at both 395 and 318 nm. While the low-
energy excitation wavelength excites the far red edge of the
absorption bands (see parts D and E of Figure 1) and little excess
energy is delivered to the pigment, the high-energy 318-nm light
excites the pigment in the high-energy wing of the ground-state
-
1
absorption band and introduces 6500 cm
of additional
excitation energy compared to the 395-nm pump pulses. Our
setup allows us to perform the two PP experiments nearly
simultaneously with ca. 10 ms time intervals between different
pump pulses47 so that the artifacts due to sample degradation
or long-term laser stability are essentially excluded. The
normalized PP traces taken at the IA and SE bands are contrasted
in Figure 7. Essentially no difference is observed in photoin-
duced dynamics between the 318-nm excitation (circles and
Figure 6. (A) Time-resolved kinetic traces measured at the SE band
of FAL (closed circles) at 450 nm and TML (open squares) at 460 nm
along with the global analysis fit to the data (solid lines). (B) Sequential
SADS of FAL. (C) Sequential SADS of TML. The estimated time
constants of the spectral evolution are shown in the legends.

Photoactive Yellow Protein Chromophores
J. Phys. Chem. B, Vol. 109, No. 9, 2005 4203
Figure 7. Normalized transient absorption traces measured on FerFA
(A) and SinFA (B) with 318-nm (symbols) and 395-nm (solid lines)
excitation. The probe wavelengths are indicated next to the curves.
squares) and 395-nm excitation (solid lines) for FerFA and
SinFA. A small difference is observed in the SinFA data, where
the SE appears to decay slightly faster for the 395-nm excitation
case than the 318-nm case, but this effect can be ascribed to
different terminal values, since the yield of two-photon ioniza-
tion appears to be excitation wavelength dependent. This was
confirmed by a global analysis fit to the data that revealed near-
identical SADS and dynamics in the accessible time window
Figure 8. Dump-induced effect on the PP signals at the probe
wavelengths corresponding to SE (solid triangles), ESA (solid squares)
bands, and at the wavelengths between the GSB and SE band (hollow
circles) in SinFA (A), FAL (B), and TML (C). The probe wavelengths
are shown in the legends. Time zero is defined by the pump pulse, and
the dump pulse arrives at ca. 2.8 ps in SinFA and ca. 5 ps in FAL and
TML. The probe wavelengths are shown in the legends.
(
not shown). In the case of several near-degenerate states, one
The data in Figure 8 show that both the SE and ESA bands
are instantaneously diminished by the applied dump pulses
negative ∆∆OD signal indicates a decrease in PP signal and
would expect an additional phase in the dynamics observed upon
blue-wing excitation (which would reflect the relaxation between
these states). Since the observed dynamics is nearly identical,
it implies that either the relaxation among different vibrational
substates (or, potentially, different electronic states) does not
occur or that it is much faster than the time resolution of the
experiment (370 fs in the case of 318-nm excitation) allows to
detect.
(
positive means increase), while at the wavelength between the
GSB and SE (open circles), the dump effect exhibits a slower
dynamics, with a lifetime of ca. 300 fs in SinFA and ca. 1.5 ps
in FAL and TML. This dynamics is due to a short-lived
unrelaxed ground state that is populated by the dump pulse.3
When the molecule decays to the ground state via either
fluorescence or internal conversion (processes much slower than
the action of dump pulse), this short-lived ground-state inter-
mediate never accumulates enough population to be observed
directly in the PP experiment. However, for TMpCA and
OMpCA at pH ) 10.5, this species is clearly distinguishable
in the PP signals alone, and the possible reasons for this are
discussed below. As in similar phenolate-containing photo-
receptor molecules, the terminal PP signals in these systems
are unaffected by the presence and timing of the dump pulse
6,45,49
3
.3. PDP: SinFA vs TML and FAL. In the previous
investigations on TMpCA, PDP spectroscopy was particularly
insightful in characterizing and identifying overlapping bands
in the measured PP spectra and elucidating underlying pathways
of the photoinduced dynamics.3
6,37,45,48
In TMpCA, the IA at
the wavelengths between the GSB and SE was attributed to an
unrealaxed ground state of the molecule.36 Similar PDP experi-
ments were performed here on SinFA, FAL, and TML at pH )
1
0.5 (excited using 395-nm light; the NOPA could be used to
generate 505-nm dump pulses). The dump-induced effects on
the PP signals are shown in Figure 8, where the PP signals
measured in the SE and ESA bands (Figures 6 and 5F) as well
as in the wavelength region between the GSB and SE are
contrasted. The traces in Figure 8 represent double difference
absorption signal (∆∆OD), which is the difference between
pump-probe signals with and without the presence of dump
(
data not shown), which is expected for the combination of a
hydrated electron and a radical, generated via a multiphoton
3
6,37,45
mechanism.
4
. Discussion
It is clear from the absorption spectra in Figure 1, and the
3
6
pulse
time-resolved data in the subsequent figures, that chromophore
modifications and protonation state have a significant influence
on the static and dynamic spectroscopic properties of these PYP
model chromophores. In particular, the esterfied derivatives
exhibit significantly red-shifted absorption spectra compared to
the corresponding free acids (e.g., TMpCA and OMpCA vs pCA
and TML vs. FAL); however, they are still blue-shifted
compared to the absorption spectra of the same chromophores
∆
∆OD(λ,tprobe,tdump) ) PDP(λ,tprobe,tdump) - PP(λ,tprobe) -
DP(λ,tprobe - tdump) (1)
PDP(λ,tprobe,tdump) denotes the PP signal with the dump pulse
present, PP(λ,tprobe) is PP signal without the dump pulse (i.e.,
the “normal” PP signal), and DP(λ,tprobe - tdump) is a DP signal
3
2
(
i.e., pump pulse is off and dump pulse is present). In our case,
bound to the PYP apoprotein. This is indicative of the special
interactions between the protein environment and the embedded
where the wavelength of the dump pulse is nonresonant with
the absorption of the sample, this signal accounts only for the
scattering and cross-phase modulation due to the dump pulse.
50
chromophore. This esterfication-induced red-shifting is greatest
when the phenolic moiety of the pigments is deprotonated (pH

4
204 J. Phys. Chem. B, Vol. 109, No. 9, 2005
10.5). In contrast to the esterfied chromophores, the pCA
Vengris et al.
)
However, the sub-picosecond decay of ESA is almost uniform
across the whole ESA band and thus is probably caused by other
processes. The observed decay of ESA could result from
vibrational relaxation of the molecule. Different vibrational
sublevels have different Franck-Condon factors that determine
the transition probabilities. For harmonic oscillators, these
factors are a function of the force constants (curvature of PESs)
and the displacement of the potential minima on the vibrational
chromophore has three protonation states depending on whether
the carboxylic acid or the phenolic moiety is protonated or both.
It exhibits a peculiar dependence of the absorption spectrum
on the protonation state. At pH ) 2, the neutral form has its
absorption maximum at 308 nm, and upon deprotonation of the
carboxylic group (pH ) 7), the peak of the spectrum shifts to
the blue (285 nm); the deprotonation of the phenolic group at
pH ) 10.5 leads to an absorption spectrum that peaks further
to the red than the neutral form (332 nm). In the other model
chromophores, the deprotonated state at elevated pH always
exhibits the more red-shifted absorption spectrum (Figure 1),
which is a sign of the greater coupling of the charged
chromophore to the solvent.
58
coordinate. It may be that, for our pigments, the maximum
S1-S2 transition probability is achieved in a nonequilibrium
population distribution among vibrational sublevels and hence
vibrational relaxation in the S1 state results in the decrease of
ESA. Usually vibrational relaxation manifests itself as a
59
narrowing of the SE and ESA bands with a rate that is largely
6
0
This sensitivity of the protonation state to the static properties
of the chromophores indicates that the charge on the phenolic
oxygen significantly alters the electron density distribution in
the molecule and perhaps even affects the conjugation length.
This has been suggested from previous Stark measurements of
solvent independent. However, the lack of excitation wave-
length dependence of the PP data for SinFA and FerFA (Figure
7) indicates that vibrational relaxation (at least, the dynamics
observable with our time resolution) does not appreciably affect
the ESA dynamics in these pigments.
5
1
isolated PYP chromophores measured at different pH values
Another explanation of this peculiar behavior of the ESA vs
SE might arise from different closely spaced electronic states
that can be accessed by 310 nm excitation. Several near-
degenerate states have been predicted for pCA in the protein
and from ab initio calculations on a model chromophore from
the green fluorescent protein, which has a structure related to
the PYP chromophores studied here.52 In the locked pigments,
a significant narrowing of the absorption spectra is observed
compared to unlocked derivatives. One explanation for this may
be the fact that, in these pigments, the ground-state distribution
of torsional angles across the C1sCR and the CRdCâ bonds
cannot occur and does not contribute to the inhomogeneous
broadening of the spectra. Alternatively, the electronic state
structure of the chromophore (i.e., the bands due to closely
spaced excited states) could have been partially modified by
added bonds. In pCA, quantum calculations have predicted that
4
6
environment by quantum-chemical calculations and were
5
4
indeed observed in gas-phase experiments. When this mixture
of different electronic states with different ESA and SE cross
sections is excited, the relaxation between these states can give
rise to a complex behavior of ESA and SE bands that may be
similar to the one observed here. However, also in this case, a
dependence on the excitation wavelength should be expected,
because different excitation wavelengths would presumably
excite different mixtures of electronic states which, in turn,
would result in different relaxation dynamics. The observed
absence of the excitation wavelength dependence implies that
multiple electronic states responsible for the lowest absorption
band of PYP, as predicted by quantum-chemical calculations,46
either do not exist (i.e., there is only one electronic state
responsible for the lowest-lying absorption band) or, more likely,
interconvert into each other with a rate faster than our time
resolution (370 fs in the case of 318-nm excitation) allows to
detect. It must be noted though that the quantum-chemical
calculations were made on the pigment different from the ones
on which the excitation wavelength dependence was measured.
multiple excited electronic states overlap, which gives rise to a
broad absorption spectrum.4
6,53
Levy and co-workers have
5
4
observed effects of exciting into these states in the gas phase;
however, direct observation of these states in condensed phases
and their influence in ensuing photochemistry has yet to be
established.
4
.1. Excited-State Dynamics. All the investigated chromo-
phores exhibit a dynamic red-shift of the SE band on a time
scale ranging from 200 fs (FAL and TML, Figure 6) to 2.8 ps
(
(
SinFA, Figure 5). Where time resolution allows its detection
e.g., 395-nm excitation datasets), the shifting dynamics is
clearly multiexponential (Figure 6) with a sub-100-fs component
and several slower phases extending to ca. 2 ps. In many of
these chromophores, little or no decay of the emission is
observed on these time scales (i.e., the amplitude of the SE band
remains largely unchanged). In the previous studies of pCA at
pH ) 10.5, this spectral shift was not observed, possibly due
to a poorer time resolution.33,34 Both solvation dynamics and
In comparison with the locked chromophores (TML and
FAL), which remain in the excited state for several nanoseconds,
all the studied chromophores exhibit dramatically shortened
excited-state lifetimes (Table 1). This rapid quenching, which
1
0,61,62
is typical of isomerizing chromophores,
is attributed to
an enhanced internal conversion induced by the motion in the
twisting coumaryl backbone of the chromophore.3
3,63
Upon
twisting are manifested as a dynamic red-shifting of fluorescence
or SE bands.1
0,55-57
A comparison of the dynamics observed in
excitation, a rotation around a single or double bond, or a “hula
twist”, occurs, which simultaneously decreases the S1 energy
and increases that of S0, which leads to rapid internal conver-
6
4
the locked vs free chromophores can be used to separate these
effects.
3
5,38
6
5
sion. After relaxation to the ground state, the spatial config-
uration of the chromophore (and the solvent) is no longer in
the minimum energy state and the system rearranges to relax
into a lower energy configuration on the ground-state PES.
Depending on the shape of the ground- and excited-state PESs,
this process may result in the formation of the (meta)-stable
isomers or a transient ground state species that quickly returns
back to the original ground state (Figure 9). Often conical
intersections (true intersecting electronic states) are involved
In a number of the investigated chromophores, a decay of
the ESA band is observed on the same time scale (500 fs to
2
contrast, the amplitude of the SE band does not decrease
appreciably on these time scales. Such ESA decay cannot
directly be attributed to solvation because the latter primarily
manifests itself as a band shift rather than a change in the
amplitude of the bands. Sometimes, solvation may account for
amplitude changes in the spectral region where ESA and SE
bands overlap, if the two bands shift differently upon solvation.
ps) as the solvation-induced SE shift (Figures 2-5). In
6
6
in very fast (∼1 ps) internal conversion rates and may

Photoactive Yellow Protein Chromophores
J. Phys. Chem. B, Vol. 109, No. 9, 2005 4205
TABLE 1: Summary of the Excited-State Lifetimes of the
Studied Pigmentsa
charge movement upon excitation is a crucial factor determining
the isomerization pathway of the PYP chromophores, it becomes
clear why, in the PYP protein system, the embedded chro-
mophore is deprotonated (see below, section 4.3).
pigment
pCA
pH
lifetime (ps)
2
2.6
7
1
7
1
7
1
7
1
7
1
2.7
9.4
3.7
2
2.1
2.5
19
Among the unlocked and unsubstituted chromophores, only
the pCA at pH)10.5 is doubly charged; other pigments have
neutral ester groups on the coumaryl tail that are more likely to
accommodate an electron density shifted from the phenolic
oxygen. Indeed, the lifetime of pCA at pH ) 10.5 is closer to
that of FerFA and SinFA. Because the typical excited-state
lifetimes observed for all the unsubstituted, unlocked model PYP
chromophores are ∼2-3 ps and the excited-state lifetime for
0.5
0.5
0.5
0.5
0.5
TMpCA
OMpCA
FerFA
8.2 (20% 44 ps)
26
18
3800
2100
SinFA
2
2,23
native PYP is ∼2 ps,
the pCA chromophore at pH ) 10.5
FAL
TML
10.5
10.5
is less relevant for understanding the excited-state dynamics of
PYP.
a
The lifetime is defined as the inverse rate of the step in the
evolutionary model (see Experimental Section), during which the largest
fraction of stimulated emission is lost.
The excited-state lifetime is also significantly longer in the
pigments with the oxymethyl substitutions on the phenolic ring
(FerFA and SinFA). It is tempting to ascribe this effect to the
increased bulkiness of the phenolic end of the molecule, where,
upon isomerization, the molecule would have to “drag along”
a greater mass and experience a greater resistance from the
solvent. However, because the carboxylic end of the molecule
is much lighter than the phenolic ring (even without the
oxymethyl substitutions), it is natural to assume that this end
would preferentially move upon the twisting of the double bond
in the coumaryl tail. These pigments are free acids and, similar
to pCA, are doubly negatively charged at pH ) 10.5. Their
excited-state lifetime is also comparable to pCA at pH ) 10.5,
which provides a hint that it is not simply the “bulkiness” of
the molecule but probably a more subtle change in the electronic
structure and charge distribution of the pigment that affect the
driving force (and hence the rate) for twisting. The pH
dependence of the SE decay time in FerFA and SinFA is
significantly weaker than (and opposite to) the one observed in
the bare pCA chromophore; the pigments with a charged
phenolic oxygen exhibit a slightly faster decay than singly
charged ones (parts A and B of Figure 5). The oxymethyl groups
are charge donor groups and should affect the dipole moment
of the pigment in the ground and excited states, which would,
in turn, affect the twisting time scales through the solvent by
enlarging or reducing the number of coordinated solvent
molecules and thus increasing or reducing the friction for
twisting. Preliminary experiments exploring the dependence of
SE decay times on the viscosity of the solvent (data not shown)
have confirmed that increasing the viscosity does slow the
excited-state lifetime.
Figure 9. Model PESs along the twisting coordinate of the pigment.
(A) Barrierless evolution in the ground state, resulting in the broad
unrelaxed ground-state absorption; (B) evolution involving the ground-
state barrier, resulting in the sharp ground-state absorption.
participate in the dynamics observed in the photoreceptors
6
7-69
studied here.
Apart from solvated electrons, no long-lived photoproducts
which could be attributed to the isomerized molecules) are
(
observed in any of the chromophores studied, which means that
the full trans-cis isomerization does not occur in the pigments
in solution or occurs with a yield that is lower than our resolution
allows us to detect (ca. 3%). The ultrafast quenching of the
excited state results from the twisting reaction when the
chromophore “attempts” to isomerize and twists in the excited
state; however, after the intersystem crossing to the ground state,
the shape of the PES is such that the molecule simply relaxes
back to the original all-trans conformation (Figure 9A). For
TmpCA, this behavior is fully consistent with the predictions
based on quantum mechanical/molecular mechanics calculations
The characterization of the PESs involved is an important
component in understanding the photophysics underlying the
photoinduced twisting reactions, because these surfaces ulti-
mately determine the driving force for the excited- and ground-
state evolution. An intriguing question about the model PYP
chromophore is whether an energy barrier is hindering the
excited-state twisting process. This has been discussed exten-
67
by Groenhof et al.
The unlocked chromophores without substitutions on the
phenolic ring exhibit the lifetimes in the range between 2 and
3
.7 ps, with the exception of pCA at pH ) 10.5, which exhibits
sively for the isomerization of the chromophore in bacterio-
33
48,61,70,71
a 9.4-ps lifetime (Table 1). A possible explanation as to why
the excited-state lifetime is prolonged so much by the depro-
tonation of the phenolic oxygen in this particular pigment could
be that it results in a different charge redistribution in the
molecule upon excitation, and hence in a different excited-state
pathway, the molecule follows after the excitation. Electro-optic
measurements show that the magnitude of the charge displace-
ment upon excitation of TMpCA and TML is significantly
reduced when measured at neutral pH, where the phenolic
moiety of the pigments is protonated.5 If one assumes that the
rhodopsin,
where a three-state model, which includes
a small excited-state energy barrier, was favored over a two-
state barrierless model. The dual-excitation wavelength PP
experiment provides evidence that no significant energy barrier
exists for twisting in the excited state for the FerFA and SinFA
pigments: the observed quenching dynamics is identical regard-
less of the excitation wavelength (310 nm vs 395 nm). If the
excess energy of 310-nm excitation contributed to crossing an
excited-state barrier, the observed twisting would be faster and/
or occur with a higher yield than using a more red excitation
1

4
206 J. Phys. Chem. B, Vol. 109, No. 9, 2005
Vengris et al.
wavelength, which is obviously not the case. In contrast, for
PYP, an excited-state energy barrier was proposed to explain
both steady-state72,73 and ultrafast experiments. It must be
noted though that the pigments studied with dual PP spectros-
copy (FerFA and SinFA at pH ) 10.5) exhibit relatively long
excited-state quenching times compared to those of solvation
and/or vibrational relaxation. Similar experiment on TMpCA
phenomena occur on <2-ps time scales, similar to the ones
observed here, and are responsible for the dynamics observed
in FAL and TML.
38
The fact is that the unrelaxed ground-state species is so
pronounced (with well-defined separate spectral band with its
own dynamics) in TMpCA and OMpCA at pH ) 10.5 but is
difficult to detect (a slight shoulder in the excited-state spectrum)
in other pigments deserves comment. Even in TMpCA and
OMpCA at neutral pH, only hints (the shoulder on the red part
of the ESA band) for the formation of this species are observed.
SinFA and FerFA show a similar shoulder in the spectrum at
pH ) 10.5 (parts E and F of Figure 5) but not at pH ) 7 (parts
C and D of Figure 5). Several factors may influence the intensity
of this feature: (1) in some cases, the lifetime of the unrelaxed
ground state can be quite short (compared to the lifetime of the
excited state), and the population of this state is never enough
to be able to observe it in PP signals; (2) its transition dipole is
weak, and its induced absorption is dominated by other
overlapping bands; (3) the ground-state absorption will be
dominated by the SE and ESA bands. The first two factors can
play a role only if the observed ground-state intermediate is
due to an inherent structural rearrangement of the molecule;
because solvation/vibrational relaxation time scales do not
strongly depend on the solute, neither do these processes affect
the transition dipole moment of the pigments. At the same time,
the excited-state lifetimes of TMpCA and OMpCA at pH ) 7
are not significantly different from those at pH ) 10.5
(Table 1), suggesting that the same amount of this ground-state
intermediate would be formed, if it was due to the ground-state
solvation or vibrational relaxation. This allows us to conclude
that the observed ground-state band and its dynamics in TMpCA
and OMpCA at pH ) 10.5 are the result of a specific pathway
(currently planned), where the twisting rate is much faster and
commensurate with the solvation/vibrational relaxation time
scales, would allow us to establish whether the twisting time
scales of this pigment in solution are sensitive to the excitation
wavelength, in other words, whether it involves an excited-state
energy barrier as predicted by Groenhof et al.74
4
.2. Ground-State Dynamics. In TMpCA and OMpCA at
pH ) 10.5, an IA band between the GSB and SE bands (parts
E and F of Figure 4) is clearly observed with a delay after the
excitation. For TMpCA at pH ) 10.5, this band has been studied
in detail with PDP spectroscopy and was attributed to a
structurally unrelaxed ground-state intermediate of the pig-
63
ment. Moreover, a similar ground-state intermediate has been
shown to occur in PYP as well, which highlights the similarity
37
of dynamics of TMpCA and PYP. A similar band was also
observed in another PYP model chromophore containing a
3
4
thiophenyl group, the chromophore of the green fluorescent
protein, (with a similar conjugated π-electron system as the
pigments investigated here) and several molecules that undergo
twisting-induced charge-transfer reactions.4 This, along with
the fact that this band is observed in OMpCA as well (albeit
slightly obscured by the pump scatter at 395 nm, Figure 4E),
indicates that the observation of this band is not unique to the
presence of the thioester group.
5,75
The PDP results on SinFA, TML, and FAL (Figure 8) show
that an unequilibrated ground-state species is also observed in
these pigments, but it is much less pronounced than in TMpCA.
In SinFA, instead of a clearly separated IA band, it manifests
itself in pump-probe signals as a shoulder in the red part of
the ESA band (18-ps SADS in Figure 5F). The late excited-
state SADS of many other pigments exhibit a similar feature
(only occurring in these pigments) in the excited-state relaxation,
whereas in other pigments (including the locked ones), the
observed spectral features are ascribed to ground-state solvation/
vibrational relaxation.
In the case of a structural reorganization, the precise dynamics
will be a function of the ground- and excited-state PES along
the twisting coordinate (Figure 9). The lifetime of the ground-
state intermediate is directly governed by the slope of the
ground-state PES (the driving force for the return to the
equilibrated ground state). To a first approximation, the transi-
tion dipole at a given point on the PES should be the same for
the IA and SE; however, depending on the shape of the PES,
the population distributions can be different. For example, if
the ground-state PES has a local minimum, a significant amount
of the population will be localized there for some time and when
projected on an excited state PES, will give a rise to a more-
defined band (Figure 9B). However, if the ground-state potential
is shallow and monotonic, a broad distribution of the population
will occur and the resulting IA will be a relatively featureless
broad band that is easily masked by other spectral features
(Figures 3 (parts B and D), 4 (parts C and D), and 5 (parts E
and F)) which is the manifestation of ground-state intermediates.
In all these pigments, the transients taken around the zero-
crossing points of the spectra show similar behavior to that
observed in pCA at pH ) 10.5 (Figure 2B), where at early times
the PP signal is dominated by the SE, and as the latter shifts to
the red due to solvation (and perhaps twisting), the ground-
state intermediate absorption rises and then decays.
The unrelaxed ground state is also present in the locked
pigments, which have much longer excited-state lifetimes and
cannot twist, although here it is completely masked by the
excited-state spectral features, and only PDP experiments can
clearly uncover it (parts B and C of Figure 8). The fact that the
pigments without rotational degrees of freedom also feature this
ground-state species indicates that it is not a unique signature
of a twisting reaction and must result from other processes such
(
Figure 9A). Because PDP spectroscopy, particularly in com-
36,37,45
bination with global analysis,
simultaneously explores both
3
6
the ground- and excited-state dynamics, the corresponding PES
as ground-state solvation and/or vibrational relaxation. When
the pigments are dumped from the excited state, their dipole
moment and polarizability changes; the configuration of the
surrounding solvent molecules, adjusted to accommodate the
excited state pigment, no longer corresponds to the energetic
minimum. The molecule might also be dumped into a vibra-
tionally excited ground state and have some excess vibrational
energy that needs to be dissipated in the solvent. Thus, after
dumping, the reaction akin to the excited-state solvation or
7
5
can be reconstructed from the data.
4.3. Relationship to PYP. The photocycle dynamics of PYP
has the most in common with the OMpCA and TMpCA
2
1,24,37
chromophores at pH ) 10.5.
These systems have similar
properties: (a) an ester group (thio- or oxymethyl) instead of a
negatively charged carboxylic oxygen on the coumaryl tail; (b)
77,78
deprotonated phenolic groups;
(c) fast excited-state quench-
ing due to twisting and/or isomerization; and (d) pronounced
unequilibrated ground-state species featuring IA at the wave-
76
vibrational relaxation will take place in the ground state. Such

Photoactive Yellow Protein Chromophores
J. Phys. Chem. B, Vol. 109, No. 9, 2005 4207
lengths between the GSB and SE.35,36 These features are
observed only with the phenolic moiety deprotonated (high pH),
while at neutral pH, their dynamics is comparable with the other
model chromophores. This strong dependence on the protonation
state indicates that a charge (re)distribution in the molecule is
the key factor that determines the dynamics of the pigment.
This has also been proposed by recent electro-optic experiments
on TMpCA, which showed a large (25 D) dipole moment
difference between the ground and excited state, making
state and the dipole moment change holds in more cases than
TMpCA. Also, a study on the different chromophores (both in
solution and in the protein) that have the rotation of the phenolic
ring locked (with the double bond free) might be insightful.
5. Conclusions
We have investigated the ultrafast dynamics of a number of
model PYP chromophores. Comparison of the free chromo-
phores with the locked chromophores allowed us to attribute
the red-shift of the SE band, occurring on the time scales of
5
1
TMpCA essentially a charge-transfer molecule at high pH,
while at the same time, the neutral form of TMpCA exhibits a
much smaller (9 D) charge displacement.51 Unfortunately, the
dipole moment change in OMpCA has not been measured. The
charge displacement effect is also noticeable on the ultrafast
dynamics of pCA: the quenching lifetime here changes from
1
00 fs to 2 ps, to predominately solvation dynamics, while
isomerization manifested itself primarily as the quenching of
stimulated emission. A transient ground-state intermediate,
resulting from unsuccessful “attempts” to isomerize can be
directly observed in PP data for TMpCA and OMpCA at pH )
1
38
2.6 ps and neutral and acidic pH to 9.4 ps at basic pH; however,
0.5. A similar intermediate has been revealed in SinFA, FAL,
the spectral features in this pigment of the excited (and, possibly,
ground) states are altered only slightly by the protonation.
Similarly long lifetimes at high pH are observed in other doubly
charged pigments: SinFA and FerFA at pH ) 10.5. In these
compounds that feature bulky (and electron-donating) groups
on the phenolic ring, the protonation effect is less obvious, most
probably because the electron density can be borrowed from
both the oxymethyl groups and the phenolic oxygen and the
charge stabilization on the phenolic oxygen by the proton
becomes less important. Ab initio calculations to verify this are
currently in progress.
and TML by the use of PDP spectroscopy. However, in these
pigments, it is attributed entirely due to ground-state solvation
and/or vibrational relaxation. Spectral features in the PP spectra
of other pigments suggest that it exists there as well. We suggest
that in TMpCA and OMpCA the ground-state intermediate
reflects the different pathway of structural evolution in excited
and ground states. The prerequisite for it is the ester group
(thiomethyl or oxymethyl) on the coumaryl tail of the pigment
and the negative charge on the phenolic oxygen. These two
features presumably enhance the charge-transfer character of
the S0fS1 transition which, through the interaction with the
environment, alters the course of twisting in these pigments. A
similar interaction with the environment may be responsible for
the photocycle initiation in PYP.
To summarize, from the experimental observations, one can
infer that the PES for twisting is significantly altered by
removing the charge from the coumaryl tail of the pigment and
replacing it with ester groups (OMpCA and TMpCA). The
strong pH dependence of the excited-state dynamics and the
Acknowledgment. M.V. is grateful to The Netherlands
Organization of Fundamental Research of Matter; D.S.L. would
like to thank the Human Frontier Science Program Organization
for providing financial support with a long-term fellowship.
51
recent Stark experiments suggest that the twisting is probably
stabilized, or even funneled, by the solvent (or protein)
environment that reacts to the large change in dipole moment
upon excitation that is characteristic to these particular chro-
mophores. The short excited-state lifetimes of all nonlocked
pigments suggest that all of them twist thereby dissipating the
excitation energy, but only TMpCA and OMpCA at pH ) 10.5
twist differently, resulting in an unequilibrated ground state also
References and Notes
(1) Feringa, B. L.; van Delden, R. A.; Koumura, N.; Geertsema, E.
M. Chem. ReV. 2000, 100, 1789.
(
(
(
2) Hampp, N. Chem. ReV. 2000, 100, 1755.
3) Irie, M. Chem. ReV. 2000, 100, 1685.
3
6,37
observed in PYP.
Perhaps the same twisting pathway leads
4) Kochendoerfer, G. G.; Mathies, R. A. Isr. J. Chem. 1995, 35, 211.
to the initiation of PYP photocycle: the photocycle is initiated
as a response of the protein to the changed dipole moment which
affects the twisting of the chromophore. The hints on the precise
mechanism of such a response are offered by the recent visible
pump-IR probe experiments on wild-type PYP and the mutants,
which showed that the hydrogen bond between the phenolate
oxygen of the chromophore and glutamate-46 residue in the
protein was instantaneously weakened upon excitation, presum-
ably due to a charge redistribution on the pigment. Such a
change in the hydrogen bonding network around the pigment
can then provide a “feedback” for the excited-state dynamics,
i.e., funnel the excited-state twisting of the chromophore,
ultimately resulting in the later photocycle intermediates. This
hypothesis is also in line with recent visible PP and PDP
(5) Oesterhelt, D.; Stoeckenius, W. Nat. New Biol. 1971, 233, 149.
6) Kennis, J. T. M.; Crosson, S.; Gauden, M.; van Stokkum, I. H.
M.; Moffat, K.; van Grondelle, R. Biochemistry 2003, 42, 3385.
7) Hellingwerf, K. J.; Hendriks, J.; Gensch, T. J. Phys. Chem. A 2003,
107, 1082.
(
(
(8) Sumitani, M.; Nakashima, N.; Yoshihara, K.; Nagakura, S. Chem.
Phys. Lett. 1977, 51, 183.
(
9) Glasbeek, M.; Zhang, H. Chem. ReV. 2004, 104, 1929.
(10) Hamm, P.; Zurek, M.; R o¨ schinger, T.; Patselt, H.; Oesterhelt, D.;
Zinth, W. Chem. Phys. Lett. 1996, 263, 613.
(
11) Meyer, T. E. Biochim. Biophys. Acta 1985, 806, 175.
7
9
(12) Sprenger, W. W.; Hoff, W. D.; Armitage, J. P.; Hellingwerf, K. J.
J. Bacteriol. 1993, 175, 3096.
(13) Borgstahl, G. E.; Williams, D. R.; Getzoff, E. D. Biochemistry 1995,
4, 6278.
3
(14) Genick, U. K.; Borgstahl, G. E.; Ng, K.; Ren, Z.; Pradervand, C.;
Burke, P. M.; Srajer, V.; Teng, T. Y.; Schildkamp, W.; McRee, D. E.;
Moffat, K.; Getzoff, E. D. Science 1997, 275, 1471.
(15) Genick, U. K.; Devanathan, S.; Meyer, T. E.; Canestrelli, I. L.;
experiments that have shown that, in PYP, the efficiency of
Williams, E.; Cusanovich, M. A.; Tollin, G.; Getzoff, E. D. Biochemistry
initiating the photocycle decreases with time.2
5,37
If the men-
1997, 36, 8.
tioned protein response to the dipole change in the pigment is
of concerted (coherent) nature, this should be expected, because
as the time passes the vibrational coherence is lost and the
probability of entering the photocycle decreases.
(16) Genick, U. K.; Soltis, S. M.; Kuhn, P.; Canestrelli, I. L.; Getzoff,
E. D. Nature 1998, 392, 206.
17) Hoff, W. D.; Dux, P.; Hard, K.; Devreese, B.; Nugteren-Roodzant,
(
I. M.; Crielaard, W.; Boelens, R.; Kaptein, R.; van Beeumen, J.; Hellingwerf,
K. J. Biochemistry 1994, 33, 13959.
(
18) Xie, A.; Kelemen, L.; Hendriks, J.; White, B. J.; Hellingwerf, K.
J.; Hoff, W. D. Biochemistry 2001, 40, 1510.
19) van Brederode, M. E.; Hoff, W. D.; Van Stokkum, I. H.; Groot,
M. L.; Hellingwerf, K. J. Biophys. J. 1996, 71, 365.
Additional Stark experiments on these model PYP chro-
mophores at different protonation states should allow researchers
to check whether the correlation between the observed ground
(

4208 J. Phys. Chem. B, Vol. 109, No. 9, 2005
Vengris et al.
(
20) Meyer, T. E.; Tollin, G.; Causgrove, P.; Cheng, P.; Blankenship,
R. E. Biophys. J. 1991, 59, 988.
21) Baltu sˇ ka, A.; van Stokkum, I. H. M.; Kroon, A.; Monshouwer, R.;
Hellingwerf, K. J.; van Grondelle, R. Chem. Phys. Lett. 1997, 270, 263.
22) Changenet, P.; Zhang, H.; van der Meer, M. J.; Hellingwerf, K. J.;
Glasbeek, M. Chem. Phys. Lett. 1998, 282, 276.
23) Chosrowjan, H.; Mataga, N.; Nakashima, N.; Imamoto, Y.;
Tokunaga, F. Chem. Phys. Lett. 1997, 270, 267.
24) Devanathan, S.; Pacheco, A.; Ujj, L.; Cusanovich, M.; Tollin, G.;
Lin, S.; Woodbury, N. Biophys. J. 1999, 77, 1017.
25) Gensch, T.; Gradinaru, C. C.; van Stokkum, I. H. M.; Hendriks,
J.; Hellingwerf, K.; van Grondelle, R. Chem. Phys. Lett. 2002, 356, 347.
26) Chosrowjan, H.; Mataga, N.; Shibata, Y.; Imamoto, Y.; Tokunaga,
F. J. Phys. Chem. B 1998, 102, 7695.
27) Mataga, N.; Chosrowjan, H.; Shibata, Y.; Imamoto, Y.; Tokunaga,
F.; Tanaka, F. J. Lumin. 2000, 87-9, 821.
28) Mataga, N.; Chosrowjan, H.; Taniguchi, S.; Hamada, N.; Tokunaga,
F.; Imamoto, Y.; Kataoka, M. Phys. Chem. Chem. Phys. 2003, 5, 2454.
29) Mataga, N.; Chosrowjan, H.; Shibata, Y.; Imamoto, Y.; Kataoka,
M.; Tokunaga, F. Chem. Phys. Lett. 2002, 352, 220.
30) Hanada, H.; Kanematsu, Y.; Kinoshita, S.; Kumauchi, M.; Sasaki,
J.; Tokunaga, F. J. Lumin. 2001, 94, 593.
31) Cordfunke, R.; Kort, R.; Pierik, A.; Gobets, B.; Koomen, G. J.;
Verhoeven, J. W.; Hellingwerf, K. J. Proc. Natl. Acad. Sci. U.S.A. 1998,
5, 7396.
32) van der Horst, M. Structure/function relations in Photoactive Yellow
Protein. Ph.D. Thesis, University of Amsterdam, 2004.
33) Changenet-Barret, P.; Plaza, P.; Martin, M. M. Chem. Phys. Lett.
001, 336, 439.
34) Changenet-Barret, P.; Espagne, A.; Katsonis, N.; Charier, S.;
Baudin, J.-B.; Jullien, L.; Plaza, P.; Martin, M. M. Chem. Phys. Lett. 2002,
65, 285.
35) Larsen, D. S.; Vengris, M.; van Stokkum, I. H. M.; van der Horst,
M.; Cordfunke, R.; Hellingwerf, K. J.; van Grondelle, R. Chem. Phys. Lett.
003, 369, 563.
36) Larsen, D. S.; Vengris, M.; van Stokkum, I. H. M.; van der Horst,
M.; de Weerd, F. L.; Hellingwerf, K. J.; van Grondelle, R. Biophys. J. 2004,
6, 2538.
37) Larsen, D. S.; van Stokkum, I. H. M.; Vengris, M.; van der Horst,
M.; Hellingwerf, K.; van Grondelle, R. Biophys. J. 2004, 87, 1858.
38) Vengris, M.; van der Horst, M.; Zgrablic, G.; van Stokkum, I. H.
(49) Papagiannakis, E.; Larsen, D. S.; van Stokkum, I. H. M.; Vengris,
M.; Hiller, R. G.; van Grondelle, R. In preparation.
(50) Yoda, M.; Houjou, H.; Inoue, Y.; Sakurai, M. J. Phys. Chem. B
2001, 105, 9887.
(51) Premvardhan, L. L.; Buda, F.; van der Horst, M. A.; Luehrs, D.;
Hellingwerf, K. J.; van Grondelle, R. J. Phys. Chem. B 2004, 108, 5138.
(52) Voityuk, A. A.; Kummer, A. D.; Michel-Beyerle, M. E.; Rosch,
N. Chem. Phys. 2001, 269, 83.
(53) He, Z.; Martin, C. H.; Birge, R.; Freed, K. F. J. Phys. Chem. A
2000, 104, 2939.
(54) Ryan, W.; Gordon, D. J.; Levy, D. H. J. Am. Chem. Soc. 2002,
124, 6194.
(55) Jimenez, R.; Fleming, G. R.; Kumar, P. V.; Maroncelli, M. Nature
1994, 369, 471.
(56) Jordanides, X. J.; Lang, M. J.; Song, X. Y.; Fleming, G. R. J. Phys.
Chem. B 1999, 103, 7995.
(57) Kovalenko, S. A.; Ruthmann, J.; Ernsting, N. P. Chem. Phys. Lett.
1997, 271, 40.
(58) van Amerongen, H.; Valkunas, L.; van Grondelle, R. Photosynthetic
Excitons; World Scientific: Singapore, 2000.
(59) Zgrablic, G.; Vo_tchovsky, K.; Kindermann, M.; Chergui, M.;
Haacke, S. Ultrafast photophysics of the protonated Schiff base of retinal
in alcohols studied by femtosecond fluorescence up-conversion. In
Femtochemistry and Femtobiology: Ultrafast EVents in Molecular Science;
Martin, M. M., Hynes, J. T., Eds.; Elsevier: Paris, 2003.
(60) Zhang, J. P.; Chen, C. H.; Koyama, Y.; Nagae, H. J. Phys. Chem.
B 1998, 102, 1632.
(61) Gai, F.; Hasson, K. C.; McDonald, J. C.; Anfinrud, P. A. Science
1998, 279, 1886.
(62) Waldeck, D. H. Chem. ReV. 1991, 91, 415.
(63) Larsen, D. S.; Vengris, M.; van Stokkum, I. H. M.; van der Horst,
M.; Hellingwerf, K.; van Grondelle, R. Biophys. J. 2004.
(64) Liu, R. S. H.; Hammond, G. S. Proc. Natl. Acad. Sci. U.S.A. 2000,
97, 11153.
(65) Fain, B. IrreVersibilities in quantum mechanics; Kluwer Academic
Publishers: Dordrecht, Boston, London, 2000.
(
(
(
(
(
(
(
(
(
(
(
9
(
(
2
(
3
(
2
(
(66) Domcke, W.; Stock, G. AdV. Chem. Phys. 1997, 1001, 1.
(67) Groenhof, G.; Bouxin-Cademartory, M.; Hess, B.; de Visser, S.;
Brendendsen, H.; Olivucci, M.; Mark, A. E.; Robb, M. A. J. Am. Chem.
Soc. 2004, 126, 4228.
(68) Toniolo, A.; Olsen, S.; Manohar, L.; Martinez, T. J. Faraday
Discuss. 2004, 127, 149.
8
(
(
M.; Haacke, S.; Chergui, M.; Hellingwerf, K. J.; van Grondelle, R.; Larsen,
D. S. Biophys. J. 2004, 87, 1848.
(69) Ko, C.; Levine, B.; Toniolo, A.; Monohar, L.; Olsen, S.; Werner,
H.-J.; Mart ´ı nez, T. J. J. Am. Chem. Soc. 2003, 125, 12710.
(70) Gonzalez-Luque, R.; Garavelli, M.; Bernardi, F.; Merchan, M.;
Robb, M. A.; Olivucci, M. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 9379.
(71) Humphrey, W.; Lu, H.; Logunov, I.; Werner, H. J.; Schulten, K.
Biophys. J. 1998, 75, 1689.
(
39) Gradinaru, C. C.; van Stokkum, I. H. M.; Pascal, A. A.; van
Grondelle, R.; van Amerongen, H. J. Phys. Chem. B 2000, 104, 9330.
40) Hoff, W. D.; van Stokkum, I. H. M.; van Ramesdonk, H. J.; van
(
Brederode, M. E.; Brouwer, A. M.; Fitch, J. C.; Meyer, T. E.; van Grondelle,
R.; Hellingwerf, K. J. Biophys. J. 1994, 67, 1691.
(
41) Holzwarth, A. R. Data Analysis in time-resolved measurements.
(72) Hoff, W. D.; Kwa, S. L. S.; van Grondelle, R.; Hellingwerf, K. J.
Photochem. Photobiol. 1992, 56, 529.
In Biophysical Techniques in Photosynthesis; Amesz, J., Hoff, A. J., Eds.;
Kluwer: Dordrecht, The Netherlands, 1996.
(73) van der Horst, M. A. Unpublished results, 2004.
(74) Groenhof, G.; Bouxin-Cademartory, M.; Hess, B.; De Visser, S.
P.; Berendsen, H. J. C.; Olivucci, M.; Mark, A. E.; Robb, M. A. J. Am.
Chem. Soc. 2004, 126, 4228.
(75) Abramavicius, D.; Gulbinas, V.; Valkunas, L.; Shiu, Y.-J.; Liang,
K.; Hayashi, M.; Lin, S. J. Phys. Chem. A 2002, 106, 8864.
(76) Changenet-Barret, P.; Choma, C.; Gooding, E.; DeGrado, W.;
Hochstrasser, R. M. J. Phys. Chem. B 2000, 104, 9322.
(77) Cusanovich, M. A.; Meyer, T. E. Biochemistry 2003, 42, 4759.
(78) Kort, R.; Phillips-Jones, M. K.; van Aalten, D. M.; Haker, A.;
Hoffer, S. M.; Hellingwerf, K. J.; Crielaard, W. Biochim. Biophys. Acta
1998, 1385, 1.
(
42) van Stokkum, I. H. M.; Larsen, D. S.; van Grondelle, R. Biochim.
Biophys. Acta 2004, 1657, 82.
43) Ekvall, K.; van der Meulen, P.; Dhollande, C.; Berg, L. E.;
Pommeret, S.; Naskrecki, R.; Mialocq, J. C. J. Appl. Phys. 2000, 87, 2340.
44) Kovalenko, S. A.; Dobryakov, A. L.; Ruthmann, J.; Ernsting, N.
P. Phys. ReV. A 1999, 59, 2369.
45) Vengris, M.; van Stokkum, I. H. M.; He, X.; Bell, A.; Tonge, P.
J.; van Grondelle, R.; Larsen, D. S. J. Phys. Chem. A 2004, 108, 4587.
46) Molina, V.; Merchan, M. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
299.
47) Larsen, D. S.; Papagiannakis, E.; van Stokkum, I. H. M.; Vengris,
M.; Kennis, J. T. M.; van Grondelle, R. Chem. Phys. Lett. 2003, 381, 733.
48) Gai, F.; McDonald, J. C.; Anfinrud, P. J. Am. Chem. Soc. 1997,
19, 6201.
(
(
(
(
4
(
(79) Groot, M. L.; van Wilderen, L.; Larsen, D. S.; van der Horst, M.
A.; van Stokkum, I. H. M.; Hellingwerf, K. J.; van Grondelle, R.
Biochemistry 2003, 42, 10054.
(
1