Assignment of resonance Raman spectrum of photoactive yellow protein in its long-lived blue-shifted intermediate

Article abstract of DOI:10.1021/jp026448z

Photoactive yellow protein (PYP) is a bacterial photoreceptor containing a 4-hydroxycinnamyl chromophore. We report the resonance Raman spectra for the long-lived blue-shifted intermediate of PYP whose chromophore is isotopically labeled with ¹³C at the carbonyl carbon atom or at the ring carbon atoms. Spectra have been also measured with PYP in D₂O where the phenolic hydroxyl group of the chromophore is deuterated. All of the observed Raman bands are assigned on the basis of the observed isotope shifts and normal mode calculations using a density functional theory. The complete assignment provides a satisfactory framework for future investigations of the photocycle mechanism in PYP by vibrational spectroscopy.

Full text of DOI:10.1021/jp026448z

J. Phys. Chem. B 2003, 107, 2837-2845
2837
Assignment of Resonance Raman Spectrum of Photoactive Yellow Protein in Its Long-Lived
Blue-Shifted Intermediate
Masashi Unno,*,† Masato Kumauchi,‡ Jun Sasaki,‡ Fumio Tokunaga,‡ and Seigo Yamauchi*,†
Institute of Multidisciplinary Research for AdVanced Materials, Tohoku UniVersity, Sendai 980-8577, Japan,
and Department of Earth and Space Science, Graduate School of Science, Osaka UniVersity,
Toyonaka, Osaka 560-0043, Japan
ReceiVed: July 3, 2002; In Final Form: NoVember 15, 2002
Photoactive yellow protein (PYP) is a bacterial photoreceptor containing a 4-hydroxycinnamyl chromophore.
We report the resonance Raman spectra for the long-lived blue-shifted intermediate of PYP whose chromophore
is isotopically labeled with 13C at the carbonyl carbon atom or at the ring carbon atoms. Spectra have been
also measured with PYP in D2O where the phenolic hydroxyl group of the chromophore is deuterated. All of
the observed Raman bands are assigned on the basis of the observed isotope shifts and normal mode calculations
using a density functional theory. The complete assignment provides a satisfactory framework for future
investigations of the photocycle mechanism in PYP by vibrational spectroscopy.
Introduction
SCHEME 1: Preparation of 4-Hydroxycinnamic[9-13C1]
Acid
The photoactive yellow protein (PYP) from phototrophic
bacterium Ectothiorhodospira halophila is a small water-soluble
photoreceptor protein,1 and it has been an attractive model for
studying protein structures and dynamics. This protein has the
4-hydroxycinnamyl chromophore, which is covalently linked
to Cys69 through a thiolester bond.2,3 Photoexcitation of PYP
triggers a photocycle that involves several intermediate states.4-7
A long-lived blue-shifted intermediate denoted PYPM (also
called I2 or pB) is the putative signaling state of this photore-
ceptor protein.
SCHEME 2: Preparation of
4-Hydroxycinnamic[1,2,3,4,5,6-13C6] Acid
To understand the photocycle mechanism of PYP in atomic
details, structural characterizations of the intermediate states are
essential. Such information on the requisite time scales can be
provided by resonance Raman spectroscopy, which can probe
individual functional groups of proteins through their frequencies
and intensities. In fact, this technique has played a key role in
the study of the determination of the chromophore structure
during the photocycle.8-11 However, a completely satisfactory
framework for the interpretation of the observed spectra is not
yet available because of a lack of a complete assignment of the
chromophore vibrations.
ment of the chromophore vibrations confirms that the chro-
mophore of PYPM in solution is in the cis configuration and its
phenolic oxygen is protonated. Furthermore, the present normal
mode calculations also have an important implication for the
analysis of the Fourier transform infrared (FTIR) spectra of PYP.
With assignments in hand, vibrational spectroscopy provides a
unique approach for studying protein dynamic processes in PYP.
Previous studies9,10 reported the high-frequency region of the
resonance Raman spectra of PYPM. Here we have obtained the
spectra of PYPM that contains the isotopically labeled chro-
mophore in the 600-1700 cm-1 region. We also present the
results of normal mode calculations based on the density
functional theory (DFT). These studies allow us to assign all
of the observed Raman bands of PYPM. The complete assign-
Materials and Methods
Preparation of 13C-Labeled Chromophores. 13C-labeled
4-hydroxycinnamic acids were prepared according to the method
described previously12 with its minor modification (Schemes 1
and 2).
The 13C-labeled reagents, triethylphosphonoacetate[13C1] (Al-
drich) and 4-hydroxybenz[13C6]aldehyde (Cambridge Isotope
Laboratory), were used and were commercially available. Under
* To whom correspondence should be addressed. E-mail: unno@
tagen.tohoku.ac.jp or yamauchi@tagen.tohoku.ac.jp.
† Tohoku University.
‡ Osaka University.
10.1021/jp026448z CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/25/2003

2838 J. Phys. Chem. B, Vol. 107, No. 12, 2003
Unno et al.
a nitrogen atmosphere, 4-hydroxybenzaldehyde was treated with
phosphonate anion, which was generated from triethylphospho-
noacetate at room temperature. After 4 h stirring, ethyl 4-hy-
droxycinnamate was obtained and then hydrolyzed in 30%
potassium hydroxide. After usual workup and purification,
colorless 4-hydroxycinnamic acid was obtained. The purity was
represent a good approximation in our system. Indeed the
nonresonance Raman spectra excited at 647.1 nm of PYPdark
and an acid-induced bleached state PYPM,dark are very similar
to those of their resonance Raman spectra (not shown).
Resonance Raman Spectra of Photodegradable Samples.
We consider a system that exhibits simple A f B photode-
composition with a rate constant k, where A and B stand for
intact and decomposed species, respectively. Raman intensities
at ν cm-1 for species A and B are defined as IA(ν) and IB(ν),
respectively. Because the fraction of species A and B at time t
is e-kt and 1 - e-kt, respectively, the experimentally observed
spectrum can be expressed by
1
checked with mass spectrometry and H NMR spectroscopy.
Sample Preparations. Expression of wild-type PYP from
E. coli, chromophore reconstitution, and protein purification
were performed as described previously.13 13C-labeled PYPs
were prepared by reconstitution of apo-protein with the above-
mentioned 4-hydroxycinnamic anhydride whose carbonyl carbon
atom (C9) or ring carbon atoms (C1-C6) were labeled with
13C. These labeled samples are denoted as 13CdO and 13C6-
ring isotopomers, respectively. PYP in buffered D2O (90% D2O/
10% H2O) was prepared by proper dilution of a concentrated
protein in 100 mM citrate/200 mM phosphate buffer at pH 5.0
into D2O, and then the sample was incubated overnight at room
temperature before the measurements. PYP in D2O is designated
as an O1-D isotopomer.
Resonance Raman Spectroscopy. Resonance Raman spectra
were obtained as described previously.9,10 A liquid-nitrogen-
cooled CCD detector (Instrument S. A., Inc.) recorded the
Raman spectra after a Triax190 spectrometer (600 grooves/mm
grating, 0.19 m focal length; Instrument S. A., Inc.) removed
the excitation light, and a Spex 500M spectrometer (3600
grooves/mm grating, 0.5 m focal length) dispersed the scattered
light. The 325.0 nm line from a helium-cadmium (IK5651R-
G, Kimmon Electric Co., Ltd.) laser excited the samples at a
90° angle relative to the axis of the collection optics. A
polarization scrambler is placed at the entrance of the spec-
trometer. The depolarization ratio is defined as the intensity of
the scattered light observed with polarization perpendicular to
the polarization of the excitation light divided by the intensity
of the scattered light with polarization parallel to the incident
light. The intensities of the Raman scattered light with polariza-
tion perpendicular and parallel to that of the incident light were
measured using an ultra-violet sheet polarizer, which is placed
in front of the polarization scrambler. PYPM was produced as
a photostationary state by continuous laser illumination at 441.6
nm during the measurement. Because the probe wavelength is
t1
I(ν,t ,t ) )
e-ktIA(ν) + (1 - e-kt)IB(ν) dt
(1)
∫
0
1
t0
where the spectrum is measured between t0 and t1 and the sample
starts to expose to the laser light at t ) 0. In a slow
photodegradation limit, kt , 1, e-kt ) 1 - kt and the observed
spectrum I(ν,t0,t1) is evaluated as follows:
t1
I(ν,t ,t ) =
(1 - kt)IA(ν) + ktIB(ν) dt
∫
0
1
t0
) {t1 - t0 - 0.5k(t12 - t0 )}IA(ν) +
2
2
0.5k(t12 - t0 )IB(ν) (2)
If three spectra of I(ν,0,∆t), I(ν,∆t,2∆t), and I(ν,2∆t,3∆t) are
sequentially acquired, we can calculate the following difference
spectrum ∆I(ν,∆t), which shows the spectral changes because
of the sample decomposition:
∆I(ν,∆t) ) I(ν,∆t,2∆t) - I(ν,0,∆t) )
I(ν,2∆t,3∆t) - I(ν,∆t,2∆t)
) k∆t2[IB(ν) - IA(ν)]
(3)
Equation 3 indicates that the difference spectrum ∆I(ν,∆t) does
not change during the measurement. In this situation, the
following equation is derived:
IA(ν) ) {I(ν,0,∆t) - 0.5∆I(ν,∆t)}/∆t
(4)
only suited for the resonance enhancement of PYPM (λmax
)
This implies that the resonance Raman spectrum of the intact
species A can be obtained with eq 4, if the observed data satisfy
eq 3.
355 nm), interference from other photocycle intermediates is
absent. The measurements were made on samples contained in
a quartz spinning cell (10 mm in diameter). All spectra were
taken at room temperature (∼23 °C), and a homemade software
eliminated the noise spikes in the spectra caused by cosmic rays.
The 100 µm slit width employed in these measurements
corresponds to an instrumental resolution of ∼4 cm-1. All
Raman spectra were calibrated using neat fenchone. The peak
positions in the Raman spectra are measured by fitting the data
with Gaussian line shapes on a polynomial background using a
nonlinear least-squares fitting routine.
Results and Discussion
1. Resonance Raman Spectrum of PYPM. The time
dependence of the resonance Raman spectrum of PYPM is shown
in Figure 1A, which illustrates the gradual changes of the
prominent doublet near 1600 cm-1. This is clearly seen in the
difference spectra of I(ν,7 min,14 min) - I(ν,0 min,7 min) (trace
d) and I(ν,14 min,21 min) - I(ν,7 min,14 min) (trace e). Traces
d and e are essentially same, indicating that the present
experimental condition fulfils eq 3. Thus, the observed spectrum
can be corrected with eqs 3 and 4 to obtain the resonance Raman
spectrum of intact PYPM. We present the corrected spectra in
the rest of this paper.
Figure 1B examines the resonance Raman spectra of PYPM
as a function of the power of the probe laser light (traces g-i).
The figure also shows the spectrum of an acid-induced bleached
state PYPM,dark (trace f), which presumably contains protonated
trans-chromophore.9 Although overall spectral features do not
vary with different laser powers, there are some distinct
differences in the spectrum as evident in the difference spectrum
DFT Calculations. Among the numerous available DFT
methods, we have selected the B3LYP hybrid functional with
the 6-31G** basis set because of its high accuracy for predicting
vibrational frequencies. It has been shown that this level of DFT
calculations yield molecular force fields and vibrational frequen-
cies in excellent agreement with experiments in a variety of
systems.14 The optimized geometry, the harmonic vibrational
frequencies, and Raman intensity were calculated using the DFT
method mentioned above via the Gaussian 98 program.15 The
calculated frequencies were scaled using a factor of 0.9613. We
note that the experimental spectra are obtained under resonance
condition, but the calculated nonresonance Raman intensities

Resonance Raman Spectrum of PYP
J. Phys. Chem. B, Vol. 107, No. 12, 2003 2839
Figure 1. (A) Time dependence of the resonance Raman spectra of PYPM in 10 mM citrate/20 mM phosphate buffer pH 5.0 at (a) 0-7, (b) 7-14,
and (c) 14-21 min. The spectra were obtained with 325.0 nm excitation (5.6 mW) under continuous laser illumination at 441.6 nm (5.6 mW). The
sample concentration was 100 µM. Traces d and e are the difference spectra of spectra b minus a and spectra c minus b, respectively. (B) Resonance
Raman spectra of PYPM,dark and PYPM. Trace f: For PYPM,dark, the sample was dissolved in 20 mM citrate/HCl buffer at pH 2.0 with excitation at
325.0 nm (2.6 mW). Traces g-i: The laser power dependence of the resonance Raman spectra of PYPM in 10 mM citrate/20 mM phosphate buffer
at pH 5.0. The spectra were obtained at (g) 5.6, (h) 1.8, and (i) 0.73 mW of the 325.0 nm probe laser power. The pump laser power was 5.6 mW.
Trace j is the difference spectrum recorded at high power (5.6 mW) minus that recorded at low power (0.73 mW) where the sharp Raman band at
1163 cm-1 balance out.
(trace j) between those with high (trace g) and low (trace i)
laser powers. The most significant changes are seen at the main
bands of ∼1600 cm-1, and some minor bands appear around
1042, 1144, 1265, and 1300 cm-1. These new features in the
high power spectrum also exist in the spectrum of PYPM,dark
.
Thus, we ascribe these Raman bands to a reversible photoprod-
uct of PYPM, whose chromophore is in trans configuration like
PYPM,dark. In this study, we have performed a set of experiments
with the intermediate laser power (1.8 mW), where we can get
moderate signal intensity and relatively small fraction of the
photoproduct.
Figures 2 and 3 show the resonance Raman spectra of PYPM
in the low and high-frequency regions, respectively. We
measured the spectrum of PYPM in buffered D2O solution, where
the phenolic OH group of the chromophore is deuterated (trace
b). The resonance Raman spectra are also shown for PYP whose
chromophore is labeled with 13C at the carbonyl carbon atom
(trace c) or at ring carbon atoms (trace d). In addition, the figure
illustrates the H/D and 12C/13C difference spectra (traces e-g),
where the effect of the isotopic substitutions can be clearly seen.
The observed frequencies and isotope shifts are summarized in
Table 1. To interpret these experimental results, we next perform
normal mode calculations as described below.
Figure 2. Low-frequency resonance Raman spectra of PYPM and its
isotopomers in 10 mM citrate/20 mM phosphate buffer pH 5.0. The
spectra were obtained with 325.0 nm excitation (1.8 mW) under
continuous laser illumination at 441.6 nm (5.6 mW). The spectra for
(a) natural abundance, (b) O1-D, (c) 13CdO, and (d) 13C6-ring samples
are shown. The difference spectra of the isotope effects are also given
as traces d-g.
2. Normal Mode Calculations Based on DFT. We have
employed protonated cis-4-hydroxycinnamyl methyl thiolester
(HCMT) as a chromophore model for normal mode calcula-
tions.9,10 For geometry optimization, the starting geometry was
taken from the crystal structure of PYPM,16 and the obtained
structures are depicted in Figure 4. Table 1S (Supporting
Information) gives the optimized geometries along with the
experimental parameters of PYPM in crystal.16 There are two
possible conformations having different orientation of the
phenolic OH group (models 1 and 2). In both cases, optimized
geometry is Cs symmetry, whereas the crystal structure of
Although the energy of model 1 is ca. 1.2 kJ mol-1 higher than
that of model 2, the optimized structures are similar to each
other except for the phenolic OH moiety, e.g., the O1-C1-C2
bond angle for model 1 is 122.8°, whereas the corresponding
angle for model 2 is 117.5°. Figure 4 also shows the optimized
structure of protonated cis-HCMT whose carbonyl O2 forms a
hydrogen bond with amide nitrogen of methylamine (model 3).
16
PYPM showed a highly distorted chromophore (Table 1S).

2840 J. Phys. Chem. B, Vol. 107, No. 12, 2003
Unno et al.
observed and calculated frequencies are somewhat large (Table
1). Further calculations including protein-chromophore interac-
tions, such as steric and/or electrostatic effects of the protein
moeity, may reduce the deviations. We should also note that
the depolarization ratios of the observed Raman bands were
nearly the same (∼0.33). This implies that the observed spectra
are truly resonantly enhanced with a nondegenerate excited
state,16 and the depolarization ratio is not helpful for the
assignment of the resonance Raman spectra of PYPM.
3. Assignment. A. CdC or CdO Stretching Mode; ν11-ν15.
As depicted in Figure 3, the resonance Raman spectra from the
natural abundance (trace a) and 13CdO labeled (trace c) samples
and their difference spectrum (trace f) confirm previous
results,9,10 showing the Raman band at 1653 cm-1 for the C9d
O2 stretching mode ν11. The DFT calculations indicate that the
ν11 frequency is sensitive to the presence or absence of a
hydrogen bond at carbonyl O2, i.e., this mode undergoes a 28
cm-1 downshift in the presence of methylamine (model 1 f
3). The observed frequency (1653 cm-1) and the 13CdO shift
(-28 cm-1) are consistent with the calculated values (1660 and
-33 cm-1) of model 3, indicating the presence of the hydrogen
bond in PYPM.10
The DFT calculations show that the 1550-1650 cm-1 region
contains normal modes that are allocated to coupled C-C and
CdC stretching modes of the aromatic ring and the ethylenic
group (Figure 5). The intense Raman bands at 1599 and 1576
cm-1 have been assigned to ν13 and ν14, respectively,9,10 which
contain ring vibrations that correspond to Y8b of tyrosine.18
This assignment is confirmed by the 13C6-ring spectrum: these
Raman bands exhibit a large shift of ca. -45 cm-1 upon 13C6-
ring substitution (Figure 3), and similar isotope shifts are
predicted by the calculations (-39 ∼ -47 cm-1) (Figure 7).
The strong ν13 band has a shoulder at 1607 cm-1. We assign
this band to ν12, which can be related to Y8a of tyrosine.18
Although the ν12 band is not clear in the natural abundance
sample (traces a and b in Figure 3), this mode clearly appears
as a shoulder in the 13CdO isotopomer (trace c) and even
becomes a prominent Raman band at 1591 cm-1 for the 13C6-
ring derivative (trace d). The simulated spectra shown in Figure
7 demonstrate the same trend. The shoulder at 1515 cm-1 is
assigned to ν15, coupled symmetric C-C stretching and C-H
bending vibration of the phenolic ring (Figure 5), which is
analogous to Y19a of tyrosine.18 This mode is seen at 1479
cm-1 in the 13C6-ring spectrum, and the observed 36 cm-1
downshift coincides with the calculated shift from 1496 to 1460
cm-1 for model 3.
Figure 3. High-frequency resonance Raman spectra of PYPM and its
isotopomers. The experimental conditions and the samples are the same
as those described in Figure 2.
A recent resonance Raman study demonstrates the presence of
a hydrogen bond at carbonyl O2 for PYPM in solution.10 As
listed in Table 1S, the formation of the hydrogen bond mainly
affects the carbonyl moiety of HCMT. For example, the C9-
O2 bond distance increases by 0.007 Å, whereas the C9-S bond
distance decreases by 0.019 Å upon formation of the hydrogen
bond.
Next we have calculated vibrational frequencies as well as
Raman intensities of protonated cis-HCMT using the optimized
structures described above. On the basis of Cs symmetry, 63
fundamentals of protonated HCMT span the representation Γ
) 43 A′ + 20 A′′ vibrations. All fundamental modes are Raman
active, and A′ and A′′ type modes correspond to in-plane (νi)
and out-of-plane (γi) vibrations, respectively. Figure 5 presents
atomic displacements for selected normal modes of model 1.
In addition to the observed frequencies, Table 1 summarizes
the computed frequencies as well as approximate descriptions
of the vibrational modes. The calculated isotope shifts for the
O1-D, 13CdO, and 13C6-ring isotopomers are also shown in
the table. For the O1-D isotopomer of model 3, two hydrogen
atoms bound to the nitrogen atom of methylamine are also
replaced by deuterium atoms, because the amide groups are
expected to be deuterated in D2O. Figures 6 and 7 illustrate the
simulated Raman spectra of model 3 in four different isoto-
pomers (traces a-d) as well as the H/D and 12C/13C difference
spectra (traces e-g). Comparison of the observed (Figures 2
and 3) and calculated spectra demonstrates satisfactory agree-
ments between the experiment and calculation, allowing us to
assign all of the observed Raman bands of PYPM. Because
protonated HCMT contains 10 hydrogen atoms, ν1-ν10 cor-
respond to the C-H or O-H stretching modes. Thus, we expect
the remaining 53 normal modes in the 0-1700 cm-1 region.
The details of the assignments will be given in the following
section.
B. CdC or C-C Stretching/C-H Rocking Modes; ν17-ν23.
The DFT calculations demonstrate that the normal modes in
the 1200-1500 cm-1 region consist primarily of CdC or C-C
stretching and in-plane C-H rocking vibrations (Figure 5, Table
1). The ν17 and ν18 modes of protonated cis-HCMT involve
asymmetric stretching motions of the phenolic ring and HC7d
C8H rocking motions (B1 symmetry in local C2V) with different
phase. The observed bands at 1449 cm-1 and its shoulder at
1433 cm-1 are assigned to ν17 and ν18, respectively (Figure 3).
Because of different contributions of the ring C-C stretches,
ν18 is expected to show a larger 13C6-ring shift (-29 cm-1
)
compared to ν17 (-10 cm-1). Such shifts of -33 (ν18) and -13
cm-1 (ν17) are observed in the 13C6-ring spectrum. Upon
deuterium substitution of the phenolic OH group, the ν17 and
ν18 modes are predicted to show ∼3 cm-1 downshifts, which
are experimentally observed (-3 ∼ -7 cm-1).
We note that, although the present DFT calculations reproduce
main features of the observed spectra, there are some deviations
from the experiment. For instance, the calculations nicely predict
spectral features of the intense doublet near 1600 cm-1 (ν13 and
ν14) and its isotope shifts, but the differences between the
The 1250-1350 cm-1 region has four normal modes (ν19,
ν20, ν22, and ν23) that are combination of the aromatic ring and

Resonance Raman Spectrum of PYP
J. Phys. Chem. B, Vol. 107, No. 12, 2003 2841
TABLE 1: Observed and Calculated Vibrational Frequency (cm-1) of PYPM and Its Modelsa
νcalb model 1
νcalb model 2
νcalb model 3
νobs
assignmentd
νC9dO2
c
1688 (0, -38, -1)
1608 (0, 0, -25)
1583 (-4, -2, -41)
1552 (-7, -2, -46)
1496 (-1, 0, -36)
1433 (-3, -1, -10)
1686 (0, -38, -1)
1610 (-1, 0, -26)
1576 (-1, -2, -39)
1558 (-9, -2, -47)
1499 (-2, 0, -36)
1433 (-3, -1, -9)
1660 (0, -33, -1)
1608 (0, 0, -27)
1653 (-2, -28, 0)
1605 (0, 0, -14)
ν11
ν12
ν13
νCC(8a), νC7dC8
νC7dC8, νCC(8b)
νCC(8b), νC7dC8
νCC+δCH(19a)
δHC7dC8H(B1),
νCC+δCH(19b)
νCC+δCH(19b),
δHC7dC8H(B1)
νCC+δCH(3),
δHC7dC8H(A1)
νCC+δCH(14),
δHC7dC8H(A1)
νC1-O1(7a′),
δHC7dC8H(A1)
δHC7dC8H(A1)
νC4-C7(7a)
1582 (-4, -3, -40) 1599 (-8, -8, -46)
1551 (-7, -4, -45) 1576 (-12, -7, -44) ν14
1496 (-1, 0, -36)
1515 (-2, -2, -36)
ν15
ν17
1433 (-3, -1, -10) 1449 (-7, 0, -13)
1416 (-4, -1, -29)
1342 (-8, 0, -26)
1324 (-20, 0, -14)
1274 (-4, 0, -30)
1412 (-3, -1, -30)
1338 (-3, 0, -32)
1329 (-25, 0, -9)
1274 (-6, 0, -27)
1416 (-4, -1, -29) 1433 (-3, 0, -33)
ν18
ν19
ν20
ν22
1342 (-8, 0, -32)
1323 (-20, 0, -8)
1273 (-4, 0, -30)
1340 (-3, 0, -38)
1330 (-13, 0, -10)
1290 (-4, 0, -31)
1255 (-2, 0, -19)
1171 (0, 0, -30)
1163 (-1, 0, -4)
1154 (-247, 0, -7)
1100 (+4, 0, -14)
991 (0, -1, -29)
985 (0, -15, -1)
972 (0, 0, -1)
1254 (0, 0, -21)
1174 (-1, 0, -22)
1165 (-4, 0, -4)
1152 (-246, 0, -14)
1097 (+5, -1, -15)
991 (0, -1, -29)
985 (+1, -15, -1)
972 (0, 0, -1)
1254 (-2, 0, -19)
1171 (0, 0, -30)
1163 (-1, 0, -4)
1153 (-247, 0, -7)
1099 (+4, 0, -14)
990 (0, +2, -27)
1002 (0, -14, -2)
970 (0, -1, -1)
968 (0, 0, -8)
1266 (-3, 0, -22)
1182 (0, 0, -27)
1174 (0, 0, -8)
ν23
ν24
ν25
ν26
ν27
ν28
ν29
γ2
δCH(9a)
δCOH
δCH(15)
νCC+δCCC(18a)
νC8-C9
1002 (0, -23, 0)
980 (+3, -1, +1)
γHC7dC8H(A2)
γCH(17a)
969 (0, 0, -8)
955 (0, 0, -8)
γ4
γ5
903 (0, 0, -8)
922 (0, 0, -8)
903 (0, 0, -8)
869 (-2, -5, -4)
833 (0, 0, -11)
γCH(5)
865 (-2, -7, -4)
833 (0, 0, -11)
865 (-2, -7, -4)
821 (0, 0, -12)
891 (-1, -6, -6)
ν31
γ6
δC7C8C9
γCH(11)
820 (-5, -1, -23)
2 × 421 (0, -1, -9)
787 (0, 0, -7)
820 (-6, -1, -22)
2 × 419 (0, -1, -9)
794 (-1, 0, -7)
762 (-1, -1, -4)
706 (-3, 0, -7)
703 (0, -4, -16)
692 (-1, 0, -3)
646 (0, -11, -6)
637 (-1, 0, -20)
570 (0, -3, -6)
507 (0, -1, -11)
820 (-6, -0, -23)
2 × 421 (0, -1, -9)
788 (0, 0, -7)
855 (0, -2, -15)e
ν32
+
νCC+νCO(1)
837 (-7, -2, -19)e
2 x γ12 τCC(16a)
γ7
γ8
ν33
γ9
ν34
γ10
ν35
ν36
γ11
γCH(10a), γC4H
759 (0, -1, -5)
705 (-3, 0, -6)
703 (0, -4, -16)
692 (-1, 0, -4)
645 (0, -11, -6)
637 (-2, -1, -20)
571 (-1, -3, -7)
507 (0, -1, -11)
760 (0, -2, -5)
703 (-3, 0, -9)
705 (0, -4, -16)
686 (0, 0, -2)
648 (0, -9, -6)
637 (-2, -1, -19)
576 (-1, -3, -7)
507 (0, -1, -11)
γC8H, γCH(10a)
νCO+νCC(13)
γCO+γCH+γCC(4), γC7H
γS-C10
721 (-3, -1, -9)
663 (0, -9, -6)
641 (0, -1, -21)
γC9dO2, γC8H
δCCC(6b)
δC4C7C8
γCH+γCO+γCC(16b)
a This table summarizes the observed and calculated frequencies in the 500-1700 cm-1 region. Five normal modes that are localized in the
methyl group of HCMT are not shown. b Calculated vibrational frequencies of protonated cis-HCMT. The numbers in parentheses are the isotope
shifts of O1-D - na, 13CdO - na, and 13C6-ring - na, respectively. na ) natural abundance. c Observed vibrational frequencies of PYPM. The
numbers in parentheses are the isotope shifts of O1-D - na, 13CdO - na, and 13C6-ring - na, respectively. The numbers in italic are not accurately
determined. d The observed Raman bands are assigned to the calculated in-plane (νi) or out-of-plane (γi) normal modes of protonated cis-HCMT.
Approximate descriptions of the calculated modes are also described. For ring vibrations, the corresponding vibrational modes of tyrosine are
indicated. The (sHC7dC8Hs) moiety of the chromophore can be approximately described by the point group C2V for the two hydrogens, so that
the local symmetry description is given for the hydrogen bending modes. e Fermi resonance that is analogous to the “tyrosine doublet”.
HC7dC8H (A1) rocking motions. The ring vibration that is
similar to Y3 of tyrosine contributes to ν19, whereas ν20 contains
Y14 like ring vibrations (Figure 5). We assign these modes to
small shoulders around 1330 cm-1 (Figure 3) and tentatively
locate ν19 and ν20 at 1340 and 1330 cm-1, respectively, on the
basis of a fitting analysis. In contrast to the natural abundance
spectrum, clear bands appear at 1317 and 1301 cm-1 in the
O1-D and 13C6-ring spectra, respectively. The 1317 cm-1 band
of trace b is assigned to ν20 because of its large downshift (∼25
cm-1) upon the deutration of the phenolic OH group (Table 1).
The band at 1301 cm-1 of trace d is assigned to ν19, which
shows a large downshift (∼30 cm-1) upon 13C6-ring substitution
(Table 1).
The ν22 and ν23 are coupled vibrations between the ring
skeleton (Y7a′) and the HC7dC8H rocking motions (Figure
5). We detected well-resolved spectral features at 1290 (ν22)
and 1266 (ν23) cm-1. These features shift down and collapse
into a single broad band around 1245 cm-1 for the 13C6-ring
spectrum. A fitting analysis estimates the ν22 and ν23 frequencies
for the 13C6-ring derivative at 1259 and 1244 cm-1, respectively,
leading to a larger downshift of ν22 (-31 cm-1) compared to
that of ν23 (-22 cm-1). Similar downshifts of 31 and 19 cm-1
are seen in the simulated spectra (Figure 7) for ν22 and ν23,
respectively.
C. Aromatic Ring Modes Containing C-H Rocking Motions;
ν
24-ν28. The 1000-1200 cm-1 region contains normal modes
that involve significant contributions of the ring C-H rocking
motions. A sharp Raman band at 1174 cm-1 has been assigned
to ν25, allocated to a Y9a type18 aromatic ring motions.10 This
assignment is confirmed by a -8 cm-1 13C6-ring shift (Figure
2), which corresponds to a -4 cm-1 shift at 1163 cm-1 for
model 3 (Table 1). The present DFT calculations further predict
a presence of ν24 slightly above (8 cm-1) the intense ν25 band
(Figure 6). Because ν24 is allocated to C4-C7 stretching
coordinate (Y7a), a large isotopic shift of -30 cm-1 is expected
for the 13C6-ring derivative. The 13C6-ring spectrum of Figure
2 indeed shows a clear shoulder at 1155 cm-1, which is
assigned to ν24 of this isotopomer. We tentatively locate ν24
for the natural abundance sample at 1182 cm-1 (8 cm-1 above
the ν25 band).
Three C-H rocking vibrations of the aromatic ring (ν26, ν27,
and ν28) locate in the 1000-1150 cm-1 region (Table 1). Two
broad bands at 1042 and 1144 cm-1 are seen in Figure 3, but
these are ascribed to a photoproduct of PYPM as discussed

2842 J. Phys. Chem. B, Vol. 107, No. 12, 2003
Unno et al.
and the overtone of γ12 (-11 × 2 ) -22 cm-1) are similar,
accounting for the downshift of the whole doublet by ∼20 cm-1
(traces a f d).
Two skeleton vibrations (ν33 and ν34) are expected to locate
around 700 cm-1 (Figure 6). Although these modes mix
together, ν33 (703 cm-1 for model 3) mainly includes O1-C1
and C4-C7 stretching coordinates such as Y13 of tyrosine,18
whereas the S-C10 stretching coordinate significantly contrib-
utes to ν34 (692 cm-1 for model 3). We assign the 721 cm-1
band to ν33 on the basis of its -3 cm-1 O1-D and -9 cm-1
13C6-ring shifts (Table 1). On the other hand, there is no
candidate for ν34 in the natural abundance spectrum in Figure
2. Note that, however, the ν33 band in the 13C6-ring spectrum
has a clear shoulder at 705 cm-1, which might correspond to
ν34.
The ring deformation mode ν35, which is similar to Y6a of
tyrosine,18 can be assigned to a broad feature at 641 cm-1
(Figure 2), because the observed frequency and a 21 cm-1
downshift for the 13C6-ring isotopomer are consistent with the
calculation (637 cm-1 and -19 cm-1 shift for model 3).
E. Out-of-Plane Vibrational Modes; γ2 and γ10. There are
20 out-of-plane vibrational modes for the protonated cis-HCMT,
and two of them are observed in PYPM. The 980 cm-1 band is
assigned to the A2 type of HC7dC8H hydrogen out-of-plane
wagging mode γ2 (Figure 5) based on the calculated frequency
of ∼970 cm-1. This band is insensitive to all isotopic substitu-
tions examined here (Figure 2 and Table 1). Although the lack
of sensitivity is consistent with the assignment, a further study
using a new isotopomer such as C7-D and/or C8-D is
necessary for the firm assignment. Note that the assignment of
γ2 is consistent with the case of an 11-cis-retinal chromophore
in rhodopsin, because similar A2 type hydrogen out-of-plane
Figure 4. Optimized geometry of three models for protonated cis-4-
hydroxycinnamyl methyl thiolester.
above. Thus, no candidates are found in the 325 nm excitation
spectra of PYPM for these modes.
D. Chromophore Skeleton Modes; ν29-ν35. Below the 1050
cm-1 region, normal modes of the chromophore skeleton appear
(Figure 5). The ν29 mode involves a C8-C9 stretching motion
and has been assigned to the band at 1002 cm-1 on the basis of
its high sensitivity to 13CdO substitution (Figures 2 and 6).10
This assignment is also supported by FTIR spectra of PYPM;
mode has been detected at ∼970 cm-1 20
.
The calculated γ10 mode at 648 cm-1 for model 3 is mainly
allocated to the C9dO2 wagging motion (Figure 5). The Raman
band at 663 cm-1 is assigned to γ10 on the basis of the -9
cm-1 13CdO and -6 cm-1 13C6-ring shifts, which coincide with
the calculated shifts. The out-of-plane C-H wagging motions
at the aromatic ring and ethylenic part mix into the C9dO2
wag, accounting for the sensitivity to 13C6-ring substitution.
previous studies21,22,26,28 have found an IR band at ∼1000 cm-1
,
which shows a sensitivity to C8-D isotopic substitution.26 Note
that the 13CdO spectrum (trace c) of Figure 2 has a small band
at 1003 cm-1, which can be assigned to a ring-breathing
vibration of phenylalanine residues denoted F1.19 In fact, a
feature at 1003 cm-1 can be discerned on the relatively broad
ν29 band for all spectra displayed in Figure 2.
4. Implication. A. Chromophore Structure. Previous stud-
ies9,10 have assigned some of the Raman bands of PYPM such
as ν11, ν13, ν14, ν25, and ν29. These assignments are confirmed
in the present study, which has the benefit of a new isotopomer
13C6-ring derivative. In addition, we have successfully assigned
remaining Raman bands in the 600-1700 cm-1 region. The
complete assignment of the resonance Raman spectrum allows
us to unambiguously determine the chromophore structure of
PYPM in solution. We previously showed that the conversion
of PYPdark to PYPM leads to significant changes in the
spectrum.9,10 The most remarkable change is a ∼30 cm-1 upshift
of the ν13 band at 1558 cm-1, and the ν13/ν14 doublet appears
at 1599 and 1576 cm-1. The ν25 band also upshifts by 11 cm-1
from 1163 to 1174 cm-1. These spectral changes have been
ascribed to protonation of phenolic O1 in PYPM.8-10 In this
The DFT calculations predict the in-plane C7dC8sC9
bending mode ν31 at 869 cm-1 for model 3 (Table 1, Figure 5).
Because ring-breathing motions of the aromatic ring also
contribute to ν31, -5 and -4 cm-1 shifts are calculated for the
13CdO and 13C6-ring isotopomers, respectively. This mode is
assigned to a band at 891 cm-1 (Figure 2) on the basis of the
observed -6 cm-1 shifts for 13CdO and 13C6-ring substitutions.
A broad feature at ∼850 cm-1 (Figure 2) can be assigned as
a doublet at 837 and 855 cm-1 that arises from a Fermi
resonance between the ring-breathing vibration ν32 and the
overtone of an out-of-plane ring-bending vibration γ12 (Figure
5). Tyrosine residue exhibits a similar “tyrosine doublet” around
830-850 cm-1 18,19
The present normal-mode analysis shows
.
study, we have found that the deuteration effect on the ν32
+
that deuteration of the phenolic OH group of the chromophore
shifts down ν32 greatly (-23 cm-1) compared to γ12 (-0.3
cm-1). Thus, the separation between the ν32 and 2 x γ12
frequencies becomes too large to cause the resonance interaction
between the two energy levels effectively, and the removal of
the Fermi resonance explains the D2O-induced changes in the
band shape of the doublet (traces a f b in Figure 2). On the
2γ12 doublet around 850 cm-1 also acts as a marker for the
protonation state of the chromophore (Figure 2). This is identical
to the case of the “tyrosine doublet”, where the intensity
ratio of the doublet is a measure of the protonation or the
state of hydrogen bonding of the phenolic OH group.19 The ν33
band at 721 cm-1 also exhibits sensitivity toward O1-D
substitution (Figure 2, Table 1) because of an involvement of
other hand, the calculated 13C6-ring shifts of ν32 (-19 cm-1
)

Resonance Raman Spectrum of PYP
J. Phys. Chem. B, Vol. 107, No. 12, 2003 2843
Figure 5. Atomic displacement vectors for some vibrational modes of protonated cis-4-hydroxycinnamyl methyl thiolester (model 1 in Figure 4).
Figure 6. Simulated Raman spectra in the low-frequency region of
Figure 7. Simulated Raman spectra in the high-frequency region of
protonated cis-4-hydroxycinnamyl methyl thiolester (model 3 in Figure
protonated cis-4-hydroxycinnamyl methyl thiolester (model 3 in Figure
4). Gaussian band shapes with a 10 cm-1 width are used. The spectra
4). Gaussian band shapes with a 10 cm-1 width are used.
for (a) natural abundance and (b) O1-D, (c) 13CdO, and (d) 13C6-ring
isotopomers are shown. The difference spectra of the isotope effects
The observation of intense Raman bands (ν13, ν14) near 1600
cm-1, which are associated with vibrations along the carbon-
carbon double bond, is consistent with other cases such as
bacteriorhodopsin.21,22 The frequency of the CdC stretching
modes has been correlated with the extent of π-electron
are also given as traces d-g.
O1-C1 stretching coordinate (Figure 5). The latter observation
provides further evidence for a protonated chromophore in
PYPM.

2844 J. Phys. Chem. B, Vol. 107, No. 12, 2003
Unno et al.
delocalization of the retinal chromophore.21 In the case of PYP,
In conclusion, this study presents the resonance Raman
investigation of PYPM in solution at room temperature. We have
examined the effects of isotopic substitutions of the chro-
mophore, and all of the observed Raman bands are assigned
with the aid of DFT calculations. With these assignments in
hand, resonance Raman spectroscopy provides an important
approach for studying the photocycle mechanism in PYP.
Further studies using time-resolved resonance Raman spectros-
copy as well as complete vibrational assignments of PYPdark
and PYPL are currently in progress.
9
the frequency of ν13 and ν14 for PYPM and PYPM,dark (1575-
1600 cm-1) is significantly higher than that for PYPdark and
PYPL (∼1555 cm-1),9,10 indicating a less delocalized electronic
structure that raises the CdC bond strength in PYPM and
PYPM,dark
.
As illustrated in Figure 4, there are two stable orientations
of the phenolic OH group of the chromophore (models 1 and
2). The computed vibrational frequencies for these conforma-
tions are similar except for the intense ν13/ν14 doublet near 1600
cm-1. The splitting of the doublet is 31 cm-1 for model 1 and
18 cm-1 for model 2 (Table 1). The resonance Raman spectrum
of PYPM shows an intermediate splitting of 23 cm-1. This
observation may indicate that the chromophore in PYPM has
two configurations such as models 1 and 2. Such a disordered
structure is consistent with recent findings that the formation
of PYPM possesses global conformational changes in the
protein.23-29
Acknowledgment. We are grateful to N. Hamada (Osaka
University) for helpful discussion and K. Yoshihara (Suntory
Institute for Bioorganic Research) for assistance in preparing
the 13C-labeled compounds. This work was supported by grants
from the Naito Foundation to M.U. and by Grant-in-Aids No.
10780399 (M.U.) and No. 10044057 (S.Y.) from the Ministry
of Education, Culture, Science, Sports, and Technology of Japan.
B. Implication for FTIR Studies. A recent FTIR study30 on
PYP suggested that the vibrational mode around 1300 cm-1 is
an indicator of the trans/cis isomerization of the chromophore.
Free trans-4-hydroxycinnamic acid exhibits an IR band at 1294
cm-1, whereas this band shifts down to 1288 cm-1 upon
photoisomerization. From the analogy to 4-hydroxycinnamic
acid, a pair of bands at 1302/1286 cm-1 in the PYPdark/PYPM
difference IR spectrum was ascribed to the trans f cis
isomerization of the chromophore. However, the present study
shows that all vibrational modes around 1300 cm-1 (ν19-ν23)
are mainly allocated to the aromatic ring motions, implying a
lack of sensitivity to the structural changes around the C7dC8
bond. Our assignment is also consistent with the available FTIR
data for isotopically labeled PYP. Imamoto et al.30 showed
that the features around 1300 cm-1 for the difference FTIR
spectra are sensitive to deuterium labeling of the aromatic
ring (C2-D/C6-D) but insensitive to that of the ethylenic bond
Supporting Information Available: The table that shows
the optimized geometries along with the experimental parameters
of PYPM in crystal. This material is available free of charge
via the Internet at http://pubs.acs.org.
References and Notes
(1) Meyer, T. E. Biochim. Biophys. Acta 1985, 806, 175-183.
(2) Hoff, W. D.; Dux, P.; Devreese, B.; Roodzant-Nugteren, I. M.;
Crielaard, W.; Boelens, R.; Kaptein, R.; Van Beeumen, J.; Hellingwerf, K.
J. Biochemistry 1994, 33, 13959-13962.
(3) Baca, M.; Borgstahl, G. E. O.; Boissinot, M.; Burke, P. M.;
Williams, D. R.; Slater, K. A.; Getzoff, E. D. Biochemistry 1994, 33,
14369-14377.
(4) 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, 1691-1705.
(5) Imamoto, Y.; Kataoka, M.; Tokunaga, F. Biochemistry 1996, 35,
14047-14053.
(6) Ujj, L.; Devanathan, S.; Meyer, T. E.; Cusanovich, M. A.; Tollin,
G.; Atkinson, G. H. Biophys. J. 1998, 75, 406-412.
(7) Takeshita, K.; Imamoto, Y.; Kataoka, M.; Tokunaga, F.; Terajima,
M. Biochemistry 2002, 41, 3037-3048.
(8) Kim, M.; Mathies, R. A.; Hoff, W. D.; Hellingwerf, K. J.
Biochemistry 1995, 34, 12669-12672.
(9) Unno, M.; Kumauchi, M.; Sasaki, J.; Tokunaga, F.; Yamauchi, S.
J. Am. Chem. Soc. 2000, 122, 4233-4234.
(10) Unno, M.; Kumauchi, M.; Sasaki, J.; Tokunaga, F.; Yamauchi, S.
Biochemistry 2002, 41, 5668-5674.
(C8-D). Thus, the IR band at 1302 cm-1 can be assigned to the
31
phenolic ring stretching and C-H rocking mode ν20 of PYPdark
,
and the feature at 1286 cm-1 is tentatively assigned to ν22 (1290
cm-1) of PYPM.
We previously showed that the C8-C9 stretching mode ν29
acts as a marker for trans/cis isomerization of the chromophore.10
The trans configuration exhibits the ν29 band at ∼1050 cm-1
,
whereas ν29 for the cis form is 990-1000 cm-1. In the resonance
Raman spectrum, ν29 of PYPdark is observed as a doublet at 1042
and 1057 cm-1, and it appears as a singlet at 1054 cm-1 for
PYPdark in D2O.8-10 Previous FTIR studies24-26,30,32 showed a
corresponding doublet near 1050 cm-1 for PYPdark, and this
feature is found to be highly sensitive to the C8-D isotopic
substitution.30 Furthermore, the deuterated PYP sample shows
(11) Zhou, Y.; Ujj, L.; Meyer, T. E.; Cusanovich, M. A.; Atkinson, G.
H. J. Phys. Chem. A 2001, 105, 5719-5726.
(12) William, S. W., Jr.; William, D. E. J. Am. Chem. Soc. 1961, 83,
1733-1738.
(13) Imamoto, Y.; Ito, T.; Kataoka, M.; Tokunaga, F. FEBS Lett. 1995,
374, 157-160.
(14) Koch, W.; Holthausen, M. C. A Chemist’s Guide to Density
Functional Theory; Wiley-VCH: Weinheim, Germany, 2000.
(15) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M.
W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98,; Gaussian, Inc.: Pittsburgh,
PA, 1998.
(16) Genick, U. K.; Borgstahl, G. E. O.; 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, 1471-1475.
(17) Strommen, D. S. J. Chem. Educ. 1992, 69, 803-807.
(18) Takeuchi, H.; Watanabe, N.; Harada, I. Spectrochim. Acta 1988,
44A, 749-761.
a singlet band at 1056 cm-1 30
, indicating that the IR feature at
∼1050 cm-1 can be assigned to ν29 of PYPdark. The time-
resolved FTIR study26 has found a band at 1003 cm-1 for PYPM
in solutions, and similar band was observed at 994 cm-1 for
the frozen sample.25,30,32 This band is moderately sensitive to
C8-D isotopic substitution30 and assigned to ν29 of PYPM. The
retinal proteins exhibit some Raman bands around 1200
cm-1 21,22
and these modes are mixtures of C-C stretching and
,
C-H in-plane bending motions such as ν29 of PYP. Because
their pattern of frequencies and intensities is very sensitive to
the configuration and conformation, they are called “fingerprint
modes” of a retinal chromophore. Thus, the ν29 mode can be
designated as a fingerprint mode of a 4-hydroxycinnamyl
chromophore and is useful for both resonance Raman and FTIR
investigations of PYP.

Resonance Raman Spectrum of PYP
J. Phys. Chem. B, Vol. 107, No. 12, 2003 2845
(19) Harada, I.; Takeuchi, H. In Spectroscopy of Biological Systems;
Clark, R. J. H., Hester, R. E., Eds.; John Wiley & Sons: Chichester, U.K.,
1981; Vol. 13, chapter 3.
(20) Eyring, G.; Curry, B.; Broek, A.; Lugtenburg, J.; Mathies, R.
Biochemistry 1982, 21, 384-393.
(21) Stockburger, M.; Alshuth, T.; Oesterhelt, D.; Ga¨rtner, W. In
Spectroscopy of Biological Systems; Clark, R. J. H., Hester, R. E., Eds.;
John Wiley & Sons: Chichester, U.K., 1986; pp 483-535.
(22) Mathies, R. A.; Smith, S. O.; Palings, I. In Biological Applications
of Raman Spectroscopy Vol II; Spiro, T. G., Ed.; Wiley-Interscience: New
York, 1988; pp 59-108.
(23) Rubinstenn, G.; Vuister, G. W.; Mulder, F. A. A.; Du¨x, P. E.;
Boelens, R.; Hellingwerf, K. J.; Kaptein, R. Nat. Struct. Biol. 1998, 5, 568-
570.
(25) Kandori, H.; Iwata, T.; Hendriks, J.; Maeda, A.; Hellingwerf, K. J.
Biochemistry 2000, 39, 7902-7909.
(26) Brudler, R.; Rammelsberg, R.; Woo, T. T.; Getzoff, E. D.; Gerwert,
K. Nat. Struct. Biol. 2001, 8, 265-270.
(27) Xie, A.; Kelemen, L.; Hendriks, J.; White, B. J.; Hellingwerf, K.
J.; Hoff, W. D. Biochemistry 2001, 40, 1510-1517.
(28) Shiozawa, M.; Yoda, M.; Kamiya, N.; Asakawa, N.; Higo, J.; Inoue,
Y.; Sakurai, M. J. Am. Chem. Soc. 2001, 123, 7445-7446.
(29) Sasaki, J.; Kumauchi, M.; Hamada, N.; Oka, T.; Tokunaga, F.
Biochemistry 2002, 41, 1915-1922.
(30) Imamoto, Y.; Shirahige, Y.; Tokunaga, F.; Kinoshita, T.; Yoshihara,
K.; Kataoka, M. Biochemistry 2001, 40, 8997-9004.
(31) To be submitted for publication.
(24) Hoff, W. D.; Xie, A.; Van Stokkum, I. H. M.; Tang, X.; Gural, J.;
Kroon, A. R.; Hellingwerf, K. J. Biochemistry 1999, 38, 1009-1017.
(32) Imamoto, Y.; Mihara, K.; Hisatomi, O.; Kataoka, M.; Tokunaga,
F.; Bojkova, N.; Yoshihara, K. J. Biol Chem. 1997, 272, 12905-12908.