Article pubs.acs.org/JPCB
Vibrational Energy Flow in Photoactive Yellow Protein Revealed by Infrared Pump−Visible Probe Spectroscopy Ryosuke Nakamura* and Norio Hamada Science and Technology Entrepreneurship Laboratory, Osaka University, 2-1 Yamada-oka, Suita, Osaka 565-0871, Japan S Supporting Information *
ABSTRACT: Vibrational energy flow in the electronic ground state of photoactive yellow protein (PYP) is studied by ultrafast infrared (IR) pump− visible probe spectroscopy. Vibrational modes of the chromophore and the surrounding protein are excited with a femtosecond IR pump pulse, and the subsequent vibrational dynamics in the chromophore are selectively probed with a visible probe pulse through changes in the absorption spectrum of the chromophore. We thus obtain the vibrational energy flow with four characteristic time constants. The vibrational excitation with an IR pulse at 1340, 1420, 1500, or 1670 cm−1 results in ultrafast intramolecular vibrational redistribution (IVR) with a time constant of 0.2 ps. The vibrational modes excited through the IVR process relax to the initial ground state with a time constant of 6−8 ps in parallel with vibrational cooling with a time constant of 14 ps. In addition, upon excitation with an IR pulse at 1670 cm−1, we observe the energy flow from the protein backbone to the chromophore that occurs with a time constant of 4.2 ps.
■
INTRODUCTION Vibrational energy flow in proteins is a fundamental process that is essential to understanding how proteins function in photosensing, enzyme kinetics, and ligand binding and dissociation.1 After photoexcitation or ligand dissociation, a local conformational rearrangement around the cofactor occurs, triggering further structural changes in the protein. Simultaneously, the excess energy deposited in the cofactor is transported efficiently and rapidly toward the surrounding aqueous medium through intermediary protein motion. The propagation of conformational change and the removal of excess energy have been extensively investigated in terms of vibrational energy flow in proteins. Various time-resolved vibrational spectroscopic techniques based on a pump−probe method have been used to clarify vibrational energy flow in proteins, starting from the electronic excited state of the cofactor. The pump pulse resonant with the optical transition of the cofactor prepares the initial electronic excited state. During or after internal conversion, the subsequent probe pulse tracks the intramolecular vibrational redistribution (IVR) and vibrational cooling (VC) based on temporal changes in infrared (IR) absorption or spontaneous Raman scattering spectra. In heme proteins, the energy is transferred from the vibrationally excited state of heme to the surrounding protein with time constants of a few picoseconds and about 20 ps;2 the energy is further transferred from the protein to the aqueous medium with time constants of 6−10 ps and about 20 ps.3 Although the energy flow time constants for heme proteins have been well characterized, the initial state of the energy flow that can be prepared by a visible (Vis) pump pulse is limited to the electronic excited state. Therefore, it is © 2015 American Chemical Society
difficult to study thoroughly the anharmonic coupling between several kinds of vibrational modes. To examine the anharmonic vibrational mode coupling in proteins, two-color IR pump−probe spectroscopy and twodimensional IR spectroscopy have been widely used.4,5 The structure and dynamics of various proteins have been studied based on the frequency change and anharmonicity of vibrational modes of the protein backbone such as the amide I mode. However, the information on other vibrational modes including cofactors is usually buried in the strong absorption of the protein backbone and water molecules. In this study, we perform IR pump−Vis probe spectroscopy to clarify the vibrational energy flow of the cofactor in a protein. The pump energy dependence of the vibrational energy flow reveals the anharmonic coupling among various vibrational modes. By using a Vis probe pulse that is resonant with the electronic transition of the cofactor, the vibrational energy flow related to the cofactor can be selectively probed by the change in its absorption spectrum. Here, we study photoactive yellow protein (PYP), which is a 125-residue, 14 kDa photoreceptor protein isolated from Ectothiorhodospira halophile.6 The PYP cofactor consists of pcoumaric acid (pCA) covalently bound to Cys69 through a thioester linkage.7,8 In the ground state, pCA is in a deprotonated trans form, which is stabilized through a hydrogen-bonding network with Glu46, Tyr42, and Cys69 (Figure 1A).8 Upon photoexcitation, PYP undergoes a Received: December 30, 2014 Revised: April 20, 2015 Published: April 21, 2015 5957
DOI: 10.1021/jp512994q J. Phys. Chem. B 2015, 119, 5957−5961
Article
The Journal of Physical Chemistry B
of the signal and idler pulses from OPA in a type I AgGaS2 crystal. The IR pump pulses used in this study have peak energies of 1340, 1420, 1500, and 1670 cm−1 (Figure 1B). They correspond to the vibrational modes resonant with the IR pump pulses, which are the coupled vibration between the HCCH rocking and the ring vibration, the coupled vibration between the phenolic C−C stretching and C−H bending, the coupled vibration between vinyl bond CC stretching and the ring vibration, and the CO stretching, respectively.18 The pulse energy of the pump pulses was 1.5 μJ. The probe pulse after the sample was dispersed onto a linear image sensor with a spectrometer. The output signals were digitized and collected at the repetition rate of the laser system (1 kHz). The IR pump beam was modulated at 500 Hz by a mechanical chopper, which was frequency locked to the laser pulse train. The full width at half-maximum of the cross-correlation traces between the pump and probe pulses was 0.2 ps.
■
RESULTS Temporal Change in Transient Absorption Spectra. Figure 2 shows the transient absorption (TA) spectra at delay
Figure 1. (A) Schematic structure of the chromophore in PYP which is involved in a hydrogen-bonding network with Glu46, Tyr42, and Cys69. (B) Spectra of IR pump pulses used in this study.
photocycle with a number of intermediate states, namely, I0, I1, and I2, which involve the trans−cis isomerization of the chromophore, rearrangement of the hydrogen-bonding network surrounding the chromophore, and large structural changes in the protein.9−11 PYP returns from the I2 state to its initial ground state after a few hundred milliseconds. Vibrational dynamics in the electronic excited state of PYP have been studied by femtosecond time-resolved fluorescence spectroscopy. After photoexcitation, PYP relaxes from the Franck−Condon state to the bottom of the potential with time constants of 0.2 and 1.5 ps.12 A large part of the excess energy of a few thousands of wavenumbers is released within 0.2 ps. Interestingly, the coherent oscillations of low-frequency vibrational modes last for subpicosecond times after the vibrational relaxation.12,13 Vibrational dynamics in the electronic ground state have not yet been clarified. Pump−dump−probe spectroscopy has been used to identified the ground state spectral component, which decays with a time constant of around 4 ps;14 however, pump− dump−fluorescence spectroscopy detected a 0.2 ps decay in vibrational relaxation in the electronic ground state.15 In this study, a unified view of the vibrational dynamics in the electronic ground state of PYP is provided by IR pump−Vis probe spectroscopy.
Figure 2. Transient absorption spectra at delay times from 0.1 to 10 ps after photoexcitation of vibrational states with IR pulses centered at (A) 1420 and (B) 1670 cm−1. The stationary absorption spectrum is also shown in the top panel. Broken lines are best fits with skewed Gaussian functions. Arrows indicate peak positions of the transient absorption spectra. Dotted vertical lines are drawn at the transient absorption peak energies at 10 ps: (A) 21 230 and (B) 21 350 cm−1.
■
EXPERIMENTAL METHODS Sample Preparation. Wild-type PYP was prepared as previously reported.16,17 The sample was placed between two BaF2 windows separated by a 50 μm Teflon spacer and was moved by a homemade Lissajous scanner to prevent local heating. The optical density of the sample was 0.5 OD at 446 nm in heavy water solution. Laser Spectroscopy. The femtosecond IR pump−Vis probe spectroscopy setup was based on an amplified modelocked Ti:sapphire laser system operating at 1 kHz. The optical parametric amplifiers (OPA) was driven by 90% of the amplified pulses, and a portion of the remaining output was used to generate a broadband probe pulse with a CaF2 plate. An IR pump pulse was obtained by difference frequency generation
times from 0.1 to 10 ps after photoexcitation of vibrational states with IR pulses centered at 1420 and 1670 cm−1. The stationary absorption spectrum is also shown in the top panel. A negative signal around 440 nm is assigned to the ground state bleach of the S0 → S1 transition, whereas the positive signal located at a lower energy than the stationary absorption spectrum is assigned to the excited state absorption (ESA) from the excited vibrational levels in the S0 to the S1 state. Note that the positive peak of the raw data indicated by an arrow in Figure 2 appears at a lower energy than the original position of the ESA spectrum because of the spectral overlap with the ground state bleach signal. In this study, we analyze the TA peak energy instead of the ESA peak energy. The validity is discussed in the Supporting Information. 5958
DOI: 10.1021/jp512994q J. Phys. Chem. B 2015, 119, 5957−5961
Article
The Journal of Physical Chemistry B There are several characteristic features in the temporal changes in the TA spectra. First, the ground state bleach signal appears immediately after photoexcitation. This indicates that the vibrational excited states of pCA are predominantly generated by the IR pump pulse, not through vibrational energy transfer from other modes. Second, the peak energy of TA shifts with time to a higher energy as indicated by arrows in Figure 2. Because the peak energy corresponds to the average energy of the Franck−Condon transition from the vibrational excited state, the peak shift is interpreted as descending the vibrational ladder via IVR and VC. Finally, the signal intensity of TA decreases with time. Because the TA intensity corresponds to the population of the vibrational state and its transition dipole moment, a decrease in intensity indicates vibrational energy transfer to outside pCA or to dark states. Excitation Energy Dependence of Transient Absorption Peak Energy. The TA peak energy at each delay time was determined by fitting a skewed Gaussian function to the data. Figure 3 shows temporal profiles of the TA peak energy
Figure 4. Temporal profiles of the transient absorption peak intensity determined by the fit with a skewed Gaussian function. The temporal profiles are offset vertically for clarity. Solid lines are best fits with double- or triple-exponential functions convoluted with an instrumental response function. The broken line is an exponential decay function with a time constant of 6.6 ps.
component. The decay time is 6.6 ps. These results are listed in Table 1. The 0.2 ps rise component observed in all profiles Table 1. Best Fits for Temporal Profiles of Transient Absorption Signals excitation energy [cm−1]
Figure 3. Temporal profiles of the peak energy of the transient absorption spectra under various excitation energy conditions. Solid lines are best fits with double-exponential functions convoluted with an instrumental response function.
1340 1420 1500 1670
under various excitation energy conditions. All profiles show a clear biphasic decay. The best fit by double exponentials gives time constants of 0.2 and 14 ps, which are independent of the excitation energy. The magnitudes of the blue-shift are about 500 and 100 cm−1 at the fast and the slow decay times, respectively. The peak energy at a long delay time depends on the excitation energy. The peak energies excited at 1340, 1420, and 1500 cm−1 approach 21 280 cm−1 at a long delay, whereas the peak energy at 1670 cm−1 approaches 21 380 cm−1. Because the peak energy at a long delay time corresponds to the transition energy from the relaxed vibrational state, this result indicates that the relaxed vibrational state upon 1670 cm−1 excitation is 100 cm−1 lower than that under the other excitation conditions. Excitation Energy Dependence of Transient Absorption Signal Intensity. Figure 4 shows the temporal profiles of TA peak intensity determined by the fit with a skewed Gaussian function. We first focus on the results obtained with IR pump pulses at 1340, 1420, and 1500 cm−1. These profiles show similar behavior: they have a 0.2 ps rise component and a ∼7 ps decay component. However, for excitation at 1670 cm−1, a 4.2 ps rise component is observed in addition to the 0.2 ps rise
0.2 0.2 0.2 0.2
± ± ± ±
rise [ps]
decay [ps]
0.05 0.05 0.05 0.05, 4.2 ± 0.2
6.7 7.1 7.5 6.6
± ± ± ±
0.6 0.4 0.6 0.3
corresponds to the component observed as an ultrafast peak shift shown in Figure 3. However, neither a peak shift nor a decay corresponding to the 4.2 ps rise component is visible in time profiles at any probe wavelength. This indicates that the vibrational energy flows from the outside of pCA upon excitation at 1670 cm−1.
■
DISCUSSION The vibrational energy flow pathways based on our results are shown in Figure 5. Figure 5A shows that the excitations of the vibrational modes with IR pulses at 1340, 1420, and 1500 cm−1 suddenly relax to an identical vibrational state with a time constant of 0.2 ps, which we call the “bottleneck state” hereafter. The bottleneck state then releases energy of ∼100 cm−1 with a time constant of 14 ps, competing with relaxation to the initial ground state with a time constant of about 7 ps. For the excitation at 1670 cm−1, the energy flow dynamics is similar to that in Figure 5A. One of the differences is that the bottleneck state is located 100 cm−1 lower than that in Figure 5A and relaxes to the initial ground state with a time constant of 6.6 ps. The biggest difference between Figures 5A and 5B is 5959
DOI: 10.1021/jp512994q J. Phys. Chem. B 2015, 119, 5957−5961
Article
The Journal of Physical Chemistry B
formed from the electronic excited state on a different pathway from the photocycle. In addition, Vis pump−IR probe spectroscopy showed that GSI is likely to be a cis isomer, not a “hot” ground state (the vibrational excited state of the initial electronic ground state).21 Because our experiment does not indicate the conformation of the chromophore, we cannot distinguish GSI from a hot ground state. However, our results indicate that at least two different bottleneck states exist, which have different energies and formed from different vibrational states. We can give rough estimate of the bottleneck state energy by comparing the initial state energy prepared by an IR pulse and the amount of released energy (600−800 cm−1). It is estimated that the bottleneck states are located in the energy range of 600−900 cm−1. If the bottleneck state is a hot ground state (vibrational excited state), not a GSI, the following are candidates for the vibrational modes: skeletal modes (the inplane CC−C bending mode at 889 cm−1 or the O−C and C−C stretching mode at 766 cm−1) or the out-of-plane CO wagging mode at 652 cm−1.18 Therefore, the vibrational energy of the local modes, such as the CC stretching, the ring vibration, or the CO stretching, is efficiently transferred to the skeletal or the out-of-plane CO wagging mode. It is interesting that the bottleneck state has a lifetime quite similar to the 6−8 ps vibrational energy transfer time from a chromophore to the surrounding protein reported for heme proteins.3,22 Comparative studies on various chromophore− protein systems would hopefully shed light on universal feature of vibrational energy transfer in proteins. Third, we focus on the energy flow from the outside of the pCA observed only upon excitation at 1670 cm−1. The donor vibrational mode that is resonant with the 1670 cm−1 pulse can be attributed to the CO stretching mode of the protein backbone. We conclude that the CO stretching mode in Cys69 is the donor mode because this residue is nearest pCA (Figure 1A). It should be noted that the carbonyl CO stretching mode of Glu46, which has an absorption at 1726 cm−1 in D2O solution,23 is also a possible donor mode. However, no signal is visible in the transient absorption spectra performed by IR pump−Vis probe spectroscopy with an IR pump pulse centered at 1750 cm−1 (data not shown). Finally, the thermal effect on the energy flow dynamics observed in this study is considered. The IR pump pulses excite not only PYP but also the surrounding D2O because of the broad absorption band in the energy range. It has been reported that the vibrational energy of a cofactor is transferred to the surrounding water through protein for more than 20 ps.3 Although there is no report of the energy flow from water to the cofactor, it is reasonable that the time scale is not faster than the energy transfer from the cofactor. Therefore, in the time window of this study, the thermal effect is considered to be negligible.
Figure 5. Vibrational energy flow pathways based on our results. (A) Excitations at 1340, 1420, and 1500 cm−1. (B) Excitation at 1670 cm−1.
that the energy flow from the outside of pCA occurs with a time constant of 4.2 ps. First, we focus on the ultrafast relaxation with a time constant of 0.2 ps from the vibrational state prepared by an IR pump pulse. During the ultrafast vibrational relaxation process, energy of more than 500 cm−1 is efficiently released. The 0.2 ps vibrational relaxation in the electronic ground state has been reported by our previous pump−dump−fluorescence spectroscopy study of PYP.15 In addition, the ultrafast vibrational relaxation in the electronic excited state has also been observed by using femtosecond time-resolved fluorescence spectroscopy.12 It is essentially important to identify which modes act as an energy acceptor in the vibrational relaxation. As seen in Figure 2, the intensity of TA does not decrease during 0.2 ps vibrational relaxation. This indicates that acceptor modes in the 0.2 ps vibrational relaxation still contribute to the Franck− Condon overlap between the ground and excited electronic states. Therefore, we conclude that intramolecular vibrational modes of pCA are mainly responsible for the acceptor modes in ultrafast vibrational relaxation. It is pointed out that the efficient coupling among the vibrational modes of pCA is achieved by being incorporated into the protein because the vibrational relaxation in the electronic excited state is slowed by sitedirected mutation around the pCA or by altering the chromophore.19,20 Additional experiments on isolated pCA in solution and model PYP chromophores both in solution and in the protein might be insightful. Second, we examine the bottleneck state that releases energy of around 100 cm−1 with a time constant of 14 ps, in parallel with a decay to the initial ground state with a time constant of 6−8 ps. The 14 ps relaxation can be attributed to the VC process, where the energy is transferred to the surrounding protein. For hemeprotein, Mizutani et al. have observed biphasic decay of the excited vibrational populations with time constants of 1.9 and 16 ps.2 They attributed them to heme cooling through the collective motions of the protein and a classical diffusion process, respectively. The similarity of the time scale indicates that a classical diffusion process is responsible for the 14 ps VC process observed in this study. The decay component with a time constant of 6−8 ps corresponds to the debated component that has been reported as a new ground state intermediate (GSI) by using pump− dump−probe spectroscopy.14 It was reported that the GSI is
■
CONCLUSIONS We measured the IR pump−Vis probe spectra for PYP to reveal the vibrational energy flow related to pCA with a time resolution of 0.2 ps. We clarified the vibrational energy flow starting from the various vibrational states that were prepared by IR pump pulses. Local vibrational modes suddenly relaxed to bottleneck states with a time constant of 0.2 ps, which could be attributed to skeletal or out-of-plane wagging modes if they are not GSI but hot ground states. The bottleneck states returned to the initial ground state with time constants of 6−8 ps, 5960
DOI: 10.1021/jp512994q J. Phys. Chem. B 2015, 119, 5957−5961
Article
The Journal of Physical Chemistry B
N. A. J. The Solution Structure of a Transient Photoreceptor Intermediate: Delta 25 Photoactive Yellow Protein. Structure 2005, 13, 953−962. (12) Nakamura, R.; Hamada, N.; Ichida, H.; Tokunaga, F.; Kanematsu, Y. Coherent Oscillations in Ultrafast Fluorescence of Photoactive Yellow Protein. J. Chem. Phys. 2007, 127, 215102. (13) Mataga, N.; Chosrowjan, H.; Shibata, Y.; Imamoto, Y.; Kataoka, M.; Tokunaga, F. Ultrafast Photoinduced Reaction Dynamics of Photoactive Yellow Protein (PYP): Observation of Coherent Oscillations in the Femtosecond Fluorescence Decay Dynamics. Chem. Phys. Lett. 2002, 352, 220−225. (14) Larsen, D. S.; van Stokkum, I. H. M.; Vengris, M.; van der Horst, M. A.; de Weerd, F. L.; Hellingwerf, K. J.; van Grondelle, R. Incoherent Manipulation of the Photoactive Yellow Protein Photocycle with Dispersed Pump-Dump-Probe Spectroscopy. Biophys. J. 2004, 87, 1858−1872. (15) Nakamura, R.; Hamada, N.; Ichida, H.; Tokunaga, F.; Kanematsu, Y. Ultrafast Dynamics of Photoactive Yellow Protein via the Photoexcitation and Emission Processes. Photochem. Photobiol. 2007, 83, 397−402. (16) Imamoto, Y.; Ito, T.; Kataoka, M.; Tokunaga, F. Reconstitution Photoactive Yellow Protein from Apoprotein and p-Coumaric AcidDerivatives. FEBS Lett. 1995, 374, 157−160. (17) Mihara, K.; Hisatomi, O.; Imamoto, Y.; Kataoka, M.; Tokunaga, F. Functional Expression and Site-Directed Mutagenesis of Photoactive Yellow Protein. J. Biochem. 1997, 121, 876−880. (18) Unno, M.; Kumauchi, M.; Tokunaga, F.; Yamauchi, S. Vibrational Assignment of the 4-Hydroxycinnamyl Chromophore in Photoactive Yellow Protein. J. Phys. Chem. B 2007, 111, 2719−2726. (19) Chosrowjan, H.; Mataga, N.; Shibata, Y.; Imamoto, Y.; Tokunaga, F. Environmental Effects on the Femtosecond-Picosecond Fluorescence Dynamics of Photoactive Yellow Protein: Chromophores in Aqueous Solutions and in Protein Nanospaces Modified by SiteDirected Mutagenesis. J. Phys. Chem. B 1998, 102, 7695−7698. (20) Mataga, N.; Chosrowjan, H.; Taniguchi, S.; Hamada, N.; Tokunaga, F.; Imamoto, Y.; Katoka, M. Ultrafast Photoreactions in Protein Nanospaces as Revealed by fs Fluorescence Dynamics Measurements on Photoactive Yellow Protein and Related Systems. Phys. Chem. Chem. Phys. 2003, 5, 2454−2460. (21) van Wilderen, L. J. G. W.; van der Horst, M. A.; van Stokkum, I. H. M.; Hellingwerf, K. J.; van Grondelle, R. Ultrafast Infrared Spectroscopy Reveals a Key Step for Successful Entry into the Photocycle for Photoactive Yellow Protein. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 15050−15055. (22) Lim, M.; Jackson, T. A.; Anfinrud, P. A. Femtosecond Near-IR Absorbance Study of Photoexcited Myoglobin: Dynamics of Electronic and Thermal Relaxation. J. Phys. Chem. 1996, 100, 12043−12051. (23) Xie, A.; Hoff, W. D.; Kroon, A. R.; Hellingwerf, K. J. Glu46 Donates a Proton to the 4-Hydroxycinnamate Anion Chromophore During the Photocycle of Photoactive Yellow Protein. Biochemistry 1996, 35, 14671−14678.
simultaneously with a 14 ps VC time constant to the surrounding protein through a classical diffusion process. When the IR pump pulse was tuned to 1670 cm−1, the C O stretching modes of the protein backbone were also excited. We observed vibrational energy flow from the CO stretching of Cys69 to the vibrational mode of pCA with a time constant of 4.2 ps.
■
ASSOCIATED CONTENT
* Supporting Information S
Details of the data analysis of the transient absorption peak energies. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jp512994q.
■
AUTHOR INFORMATION
Corresponding Author
*(R.N.) Tel +81-6-6879-7755; Fax +81-6-6879-7878; e-mail r. [email protected]. Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This work was supported by JSPS KAKENHI Grant 24540323 and the Murata Science Foundation. REFERENCES
(1) Leitner, D. M. Energy Flow in Proteins. Annu. Rev. Biochem. 2008, 59, 233−259. (2) Mizutani, Y.; Kitagawa, T. Direct Observation of Cooling of Heme upon Photodissociation of Carbonmonoxy Myoglobin. Science 1997, 278, 443−446. (3) Lian, T.; Locke, B.; Kholodenko, Y.; Hochstrasser, R. M. Energy Flow from Solute to Solvent Probed by Femtosecond IR Spectroscopy: Malachite Green and Heme Protein Solutions. J. Phys. Chem. 1994, 98, 11648−11656. (4) Hamm, P.; Lim, M.; Hochstrasser, R. M. Structure of the Amide I Band of Peptides Measured by Femtosecond Nonlinear-Infrared Spectroscopy. J. Phys. Chem. B 1998, 102, 6123−6139. (5) Kim, Y. S.; Hochstrasser, R. M. Applications of 2D IR Spectroscopy to Peptides, Proteins, and Hydrogen-Bond Dynamics. J. Phys. Chem. B 2009, 113, 8231−8251. (6) Meyer, T. E. Isolation and Characterization of Soluble Cytochromes, Ferredoxins and Other Chromophoric Proteins from the Halophilic Phototrophic Bacterium Ectothiorhodospira-Halophila. Biochim. Biophys. Acta 1985, 806, 175−183. (7) Hoff, W. D.; Düx, P.; Hård, K.; Devreese, B.; Nugteren-Roodzant, I. M.; Crielaard, W.; Boelens, R.; Kaptein, R.; Van Beeumen, J.; Hellingwerf, K. J. Thiol Ester-Linked p-Coumaric Acid as a New Photoactive Prosthetic Group in a Protein with Rhodopsin-Like Photochemistry. Biochemistry 1994, 33, 13959−13962. (8) Baca, M.; Borgstahl, G. E. O.; Boissinot, M.; Burke, P. M.; Williams, D. R.; Slater, K. A.; Getzoff, E. D. Complete ChemicalStructure of Photoactive Yellow Protein - Novel Thioester-Linked 4Hydroxycinnamyl Chromophore and Photocycle Chemist. Biochemistry 1994, 33, 14369−14377. (9) Hellingwerf, K. J.; Hendriks, J.; Gensch, T. Photoactive Yellow Protein, a New Type of Photoreceptor Protein: Will This “Yellow Lab” Bring Us Where We Want to Go? J. Phys. Chem. A 2003, 107, 1082− 1094. (10) Ihee, H.; Rajagopal, S.; Srajer, V.; Pah, R.; Anderson, S.; Schmidt, M.; Schotte, F.; Anfinrud, P. A.; Wulff, M.; Keith Moffat, K. Visualizing Reaction Pathways in Photoactive Yellow Protein from Nanoseconds toSeconds. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 7145−7150. (11) Bernard, C.; Houben, K.; Derix, N. M.; Marks, D.; van der Horst, M. A.; Hellingwerf, K. J.; Boelens, R.; Kaptein, R.; van Nuland, 5961
DOI: 10.1021/jp512994q J. Phys. Chem. B 2015, 119, 5957−5961
Vibrational Energy Flow in Photoactive Yellow Protein Revealed by Infrared Pump−Visible Probe Spectroscopy Ryosuke Nakamura* and Norio Hamada Science and Technology Entrepreneurship Laboratory, Osaka University, 2-1 Yamada-oka, Suita, Osaka 565-0871, Japan S Supporting Information *
ABSTRACT: Vibrational energy flow in the electronic ground state of photoactive yellow protein (PYP) is studied by ultrafast infrared (IR) pump− visible probe spectroscopy. Vibrational modes of the chromophore and the surrounding protein are excited with a femtosecond IR pump pulse, and the subsequent vibrational dynamics in the chromophore are selectively probed with a visible probe pulse through changes in the absorption spectrum of the chromophore. We thus obtain the vibrational energy flow with four characteristic time constants. The vibrational excitation with an IR pulse at 1340, 1420, 1500, or 1670 cm−1 results in ultrafast intramolecular vibrational redistribution (IVR) with a time constant of 0.2 ps. The vibrational modes excited through the IVR process relax to the initial ground state with a time constant of 6−8 ps in parallel with vibrational cooling with a time constant of 14 ps. In addition, upon excitation with an IR pulse at 1670 cm−1, we observe the energy flow from the protein backbone to the chromophore that occurs with a time constant of 4.2 ps.
■
INTRODUCTION Vibrational energy flow in proteins is a fundamental process that is essential to understanding how proteins function in photosensing, enzyme kinetics, and ligand binding and dissociation.1 After photoexcitation or ligand dissociation, a local conformational rearrangement around the cofactor occurs, triggering further structural changes in the protein. Simultaneously, the excess energy deposited in the cofactor is transported efficiently and rapidly toward the surrounding aqueous medium through intermediary protein motion. The propagation of conformational change and the removal of excess energy have been extensively investigated in terms of vibrational energy flow in proteins. Various time-resolved vibrational spectroscopic techniques based on a pump−probe method have been used to clarify vibrational energy flow in proteins, starting from the electronic excited state of the cofactor. The pump pulse resonant with the optical transition of the cofactor prepares the initial electronic excited state. During or after internal conversion, the subsequent probe pulse tracks the intramolecular vibrational redistribution (IVR) and vibrational cooling (VC) based on temporal changes in infrared (IR) absorption or spontaneous Raman scattering spectra. In heme proteins, the energy is transferred from the vibrationally excited state of heme to the surrounding protein with time constants of a few picoseconds and about 20 ps;2 the energy is further transferred from the protein to the aqueous medium with time constants of 6−10 ps and about 20 ps.3 Although the energy flow time constants for heme proteins have been well characterized, the initial state of the energy flow that can be prepared by a visible (Vis) pump pulse is limited to the electronic excited state. Therefore, it is © 2015 American Chemical Society
difficult to study thoroughly the anharmonic coupling between several kinds of vibrational modes. To examine the anharmonic vibrational mode coupling in proteins, two-color IR pump−probe spectroscopy and twodimensional IR spectroscopy have been widely used.4,5 The structure and dynamics of various proteins have been studied based on the frequency change and anharmonicity of vibrational modes of the protein backbone such as the amide I mode. However, the information on other vibrational modes including cofactors is usually buried in the strong absorption of the protein backbone and water molecules. In this study, we perform IR pump−Vis probe spectroscopy to clarify the vibrational energy flow of the cofactor in a protein. The pump energy dependence of the vibrational energy flow reveals the anharmonic coupling among various vibrational modes. By using a Vis probe pulse that is resonant with the electronic transition of the cofactor, the vibrational energy flow related to the cofactor can be selectively probed by the change in its absorption spectrum. Here, we study photoactive yellow protein (PYP), which is a 125-residue, 14 kDa photoreceptor protein isolated from Ectothiorhodospira halophile.6 The PYP cofactor consists of pcoumaric acid (pCA) covalently bound to Cys69 through a thioester linkage.7,8 In the ground state, pCA is in a deprotonated trans form, which is stabilized through a hydrogen-bonding network with Glu46, Tyr42, and Cys69 (Figure 1A).8 Upon photoexcitation, PYP undergoes a Received: December 30, 2014 Revised: April 20, 2015 Published: April 21, 2015 5957
DOI: 10.1021/jp512994q J. Phys. Chem. B 2015, 119, 5957−5961
Article
The Journal of Physical Chemistry B
of the signal and idler pulses from OPA in a type I AgGaS2 crystal. The IR pump pulses used in this study have peak energies of 1340, 1420, 1500, and 1670 cm−1 (Figure 1B). They correspond to the vibrational modes resonant with the IR pump pulses, which are the coupled vibration between the HCCH rocking and the ring vibration, the coupled vibration between the phenolic C−C stretching and C−H bending, the coupled vibration between vinyl bond CC stretching and the ring vibration, and the CO stretching, respectively.18 The pulse energy of the pump pulses was 1.5 μJ. The probe pulse after the sample was dispersed onto a linear image sensor with a spectrometer. The output signals were digitized and collected at the repetition rate of the laser system (1 kHz). The IR pump beam was modulated at 500 Hz by a mechanical chopper, which was frequency locked to the laser pulse train. The full width at half-maximum of the cross-correlation traces between the pump and probe pulses was 0.2 ps.
■
RESULTS Temporal Change in Transient Absorption Spectra. Figure 2 shows the transient absorption (TA) spectra at delay
Figure 1. (A) Schematic structure of the chromophore in PYP which is involved in a hydrogen-bonding network with Glu46, Tyr42, and Cys69. (B) Spectra of IR pump pulses used in this study.
photocycle with a number of intermediate states, namely, I0, I1, and I2, which involve the trans−cis isomerization of the chromophore, rearrangement of the hydrogen-bonding network surrounding the chromophore, and large structural changes in the protein.9−11 PYP returns from the I2 state to its initial ground state after a few hundred milliseconds. Vibrational dynamics in the electronic excited state of PYP have been studied by femtosecond time-resolved fluorescence spectroscopy. After photoexcitation, PYP relaxes from the Franck−Condon state to the bottom of the potential with time constants of 0.2 and 1.5 ps.12 A large part of the excess energy of a few thousands of wavenumbers is released within 0.2 ps. Interestingly, the coherent oscillations of low-frequency vibrational modes last for subpicosecond times after the vibrational relaxation.12,13 Vibrational dynamics in the electronic ground state have not yet been clarified. Pump−dump−probe spectroscopy has been used to identified the ground state spectral component, which decays with a time constant of around 4 ps;14 however, pump− dump−fluorescence spectroscopy detected a 0.2 ps decay in vibrational relaxation in the electronic ground state.15 In this study, a unified view of the vibrational dynamics in the electronic ground state of PYP is provided by IR pump−Vis probe spectroscopy.
Figure 2. Transient absorption spectra at delay times from 0.1 to 10 ps after photoexcitation of vibrational states with IR pulses centered at (A) 1420 and (B) 1670 cm−1. The stationary absorption spectrum is also shown in the top panel. Broken lines are best fits with skewed Gaussian functions. Arrows indicate peak positions of the transient absorption spectra. Dotted vertical lines are drawn at the transient absorption peak energies at 10 ps: (A) 21 230 and (B) 21 350 cm−1.
■
EXPERIMENTAL METHODS Sample Preparation. Wild-type PYP was prepared as previously reported.16,17 The sample was placed between two BaF2 windows separated by a 50 μm Teflon spacer and was moved by a homemade Lissajous scanner to prevent local heating. The optical density of the sample was 0.5 OD at 446 nm in heavy water solution. Laser Spectroscopy. The femtosecond IR pump−Vis probe spectroscopy setup was based on an amplified modelocked Ti:sapphire laser system operating at 1 kHz. The optical parametric amplifiers (OPA) was driven by 90% of the amplified pulses, and a portion of the remaining output was used to generate a broadband probe pulse with a CaF2 plate. An IR pump pulse was obtained by difference frequency generation
times from 0.1 to 10 ps after photoexcitation of vibrational states with IR pulses centered at 1420 and 1670 cm−1. The stationary absorption spectrum is also shown in the top panel. A negative signal around 440 nm is assigned to the ground state bleach of the S0 → S1 transition, whereas the positive signal located at a lower energy than the stationary absorption spectrum is assigned to the excited state absorption (ESA) from the excited vibrational levels in the S0 to the S1 state. Note that the positive peak of the raw data indicated by an arrow in Figure 2 appears at a lower energy than the original position of the ESA spectrum because of the spectral overlap with the ground state bleach signal. In this study, we analyze the TA peak energy instead of the ESA peak energy. The validity is discussed in the Supporting Information. 5958
DOI: 10.1021/jp512994q J. Phys. Chem. B 2015, 119, 5957−5961
Article
The Journal of Physical Chemistry B There are several characteristic features in the temporal changes in the TA spectra. First, the ground state bleach signal appears immediately after photoexcitation. This indicates that the vibrational excited states of pCA are predominantly generated by the IR pump pulse, not through vibrational energy transfer from other modes. Second, the peak energy of TA shifts with time to a higher energy as indicated by arrows in Figure 2. Because the peak energy corresponds to the average energy of the Franck−Condon transition from the vibrational excited state, the peak shift is interpreted as descending the vibrational ladder via IVR and VC. Finally, the signal intensity of TA decreases with time. Because the TA intensity corresponds to the population of the vibrational state and its transition dipole moment, a decrease in intensity indicates vibrational energy transfer to outside pCA or to dark states. Excitation Energy Dependence of Transient Absorption Peak Energy. The TA peak energy at each delay time was determined by fitting a skewed Gaussian function to the data. Figure 3 shows temporal profiles of the TA peak energy
Figure 4. Temporal profiles of the transient absorption peak intensity determined by the fit with a skewed Gaussian function. The temporal profiles are offset vertically for clarity. Solid lines are best fits with double- or triple-exponential functions convoluted with an instrumental response function. The broken line is an exponential decay function with a time constant of 6.6 ps.
component. The decay time is 6.6 ps. These results are listed in Table 1. The 0.2 ps rise component observed in all profiles Table 1. Best Fits for Temporal Profiles of Transient Absorption Signals excitation energy [cm−1]
Figure 3. Temporal profiles of the peak energy of the transient absorption spectra under various excitation energy conditions. Solid lines are best fits with double-exponential functions convoluted with an instrumental response function.
1340 1420 1500 1670
under various excitation energy conditions. All profiles show a clear biphasic decay. The best fit by double exponentials gives time constants of 0.2 and 14 ps, which are independent of the excitation energy. The magnitudes of the blue-shift are about 500 and 100 cm−1 at the fast and the slow decay times, respectively. The peak energy at a long delay time depends on the excitation energy. The peak energies excited at 1340, 1420, and 1500 cm−1 approach 21 280 cm−1 at a long delay, whereas the peak energy at 1670 cm−1 approaches 21 380 cm−1. Because the peak energy at a long delay time corresponds to the transition energy from the relaxed vibrational state, this result indicates that the relaxed vibrational state upon 1670 cm−1 excitation is 100 cm−1 lower than that under the other excitation conditions. Excitation Energy Dependence of Transient Absorption Signal Intensity. Figure 4 shows the temporal profiles of TA peak intensity determined by the fit with a skewed Gaussian function. We first focus on the results obtained with IR pump pulses at 1340, 1420, and 1500 cm−1. These profiles show similar behavior: they have a 0.2 ps rise component and a ∼7 ps decay component. However, for excitation at 1670 cm−1, a 4.2 ps rise component is observed in addition to the 0.2 ps rise
0.2 0.2 0.2 0.2
± ± ± ±
rise [ps]
decay [ps]
0.05 0.05 0.05 0.05, 4.2 ± 0.2
6.7 7.1 7.5 6.6
± ± ± ±
0.6 0.4 0.6 0.3
corresponds to the component observed as an ultrafast peak shift shown in Figure 3. However, neither a peak shift nor a decay corresponding to the 4.2 ps rise component is visible in time profiles at any probe wavelength. This indicates that the vibrational energy flows from the outside of pCA upon excitation at 1670 cm−1.
■
DISCUSSION The vibrational energy flow pathways based on our results are shown in Figure 5. Figure 5A shows that the excitations of the vibrational modes with IR pulses at 1340, 1420, and 1500 cm−1 suddenly relax to an identical vibrational state with a time constant of 0.2 ps, which we call the “bottleneck state” hereafter. The bottleneck state then releases energy of ∼100 cm−1 with a time constant of 14 ps, competing with relaxation to the initial ground state with a time constant of about 7 ps. For the excitation at 1670 cm−1, the energy flow dynamics is similar to that in Figure 5A. One of the differences is that the bottleneck state is located 100 cm−1 lower than that in Figure 5A and relaxes to the initial ground state with a time constant of 6.6 ps. The biggest difference between Figures 5A and 5B is 5959
DOI: 10.1021/jp512994q J. Phys. Chem. B 2015, 119, 5957−5961
Article
The Journal of Physical Chemistry B
formed from the electronic excited state on a different pathway from the photocycle. In addition, Vis pump−IR probe spectroscopy showed that GSI is likely to be a cis isomer, not a “hot” ground state (the vibrational excited state of the initial electronic ground state).21 Because our experiment does not indicate the conformation of the chromophore, we cannot distinguish GSI from a hot ground state. However, our results indicate that at least two different bottleneck states exist, which have different energies and formed from different vibrational states. We can give rough estimate of the bottleneck state energy by comparing the initial state energy prepared by an IR pulse and the amount of released energy (600−800 cm−1). It is estimated that the bottleneck states are located in the energy range of 600−900 cm−1. If the bottleneck state is a hot ground state (vibrational excited state), not a GSI, the following are candidates for the vibrational modes: skeletal modes (the inplane CC−C bending mode at 889 cm−1 or the O−C and C−C stretching mode at 766 cm−1) or the out-of-plane CO wagging mode at 652 cm−1.18 Therefore, the vibrational energy of the local modes, such as the CC stretching, the ring vibration, or the CO stretching, is efficiently transferred to the skeletal or the out-of-plane CO wagging mode. It is interesting that the bottleneck state has a lifetime quite similar to the 6−8 ps vibrational energy transfer time from a chromophore to the surrounding protein reported for heme proteins.3,22 Comparative studies on various chromophore− protein systems would hopefully shed light on universal feature of vibrational energy transfer in proteins. Third, we focus on the energy flow from the outside of the pCA observed only upon excitation at 1670 cm−1. The donor vibrational mode that is resonant with the 1670 cm−1 pulse can be attributed to the CO stretching mode of the protein backbone. We conclude that the CO stretching mode in Cys69 is the donor mode because this residue is nearest pCA (Figure 1A). It should be noted that the carbonyl CO stretching mode of Glu46, which has an absorption at 1726 cm−1 in D2O solution,23 is also a possible donor mode. However, no signal is visible in the transient absorption spectra performed by IR pump−Vis probe spectroscopy with an IR pump pulse centered at 1750 cm−1 (data not shown). Finally, the thermal effect on the energy flow dynamics observed in this study is considered. The IR pump pulses excite not only PYP but also the surrounding D2O because of the broad absorption band in the energy range. It has been reported that the vibrational energy of a cofactor is transferred to the surrounding water through protein for more than 20 ps.3 Although there is no report of the energy flow from water to the cofactor, it is reasonable that the time scale is not faster than the energy transfer from the cofactor. Therefore, in the time window of this study, the thermal effect is considered to be negligible.
Figure 5. Vibrational energy flow pathways based on our results. (A) Excitations at 1340, 1420, and 1500 cm−1. (B) Excitation at 1670 cm−1.
that the energy flow from the outside of pCA occurs with a time constant of 4.2 ps. First, we focus on the ultrafast relaxation with a time constant of 0.2 ps from the vibrational state prepared by an IR pump pulse. During the ultrafast vibrational relaxation process, energy of more than 500 cm−1 is efficiently released. The 0.2 ps vibrational relaxation in the electronic ground state has been reported by our previous pump−dump−fluorescence spectroscopy study of PYP.15 In addition, the ultrafast vibrational relaxation in the electronic excited state has also been observed by using femtosecond time-resolved fluorescence spectroscopy.12 It is essentially important to identify which modes act as an energy acceptor in the vibrational relaxation. As seen in Figure 2, the intensity of TA does not decrease during 0.2 ps vibrational relaxation. This indicates that acceptor modes in the 0.2 ps vibrational relaxation still contribute to the Franck− Condon overlap between the ground and excited electronic states. Therefore, we conclude that intramolecular vibrational modes of pCA are mainly responsible for the acceptor modes in ultrafast vibrational relaxation. It is pointed out that the efficient coupling among the vibrational modes of pCA is achieved by being incorporated into the protein because the vibrational relaxation in the electronic excited state is slowed by sitedirected mutation around the pCA or by altering the chromophore.19,20 Additional experiments on isolated pCA in solution and model PYP chromophores both in solution and in the protein might be insightful. Second, we examine the bottleneck state that releases energy of around 100 cm−1 with a time constant of 14 ps, in parallel with a decay to the initial ground state with a time constant of 6−8 ps. The 14 ps relaxation can be attributed to the VC process, where the energy is transferred to the surrounding protein. For hemeprotein, Mizutani et al. have observed biphasic decay of the excited vibrational populations with time constants of 1.9 and 16 ps.2 They attributed them to heme cooling through the collective motions of the protein and a classical diffusion process, respectively. The similarity of the time scale indicates that a classical diffusion process is responsible for the 14 ps VC process observed in this study. The decay component with a time constant of 6−8 ps corresponds to the debated component that has been reported as a new ground state intermediate (GSI) by using pump− dump−probe spectroscopy.14 It was reported that the GSI is
■
CONCLUSIONS We measured the IR pump−Vis probe spectra for PYP to reveal the vibrational energy flow related to pCA with a time resolution of 0.2 ps. We clarified the vibrational energy flow starting from the various vibrational states that were prepared by IR pump pulses. Local vibrational modes suddenly relaxed to bottleneck states with a time constant of 0.2 ps, which could be attributed to skeletal or out-of-plane wagging modes if they are not GSI but hot ground states. The bottleneck states returned to the initial ground state with time constants of 6−8 ps, 5960
DOI: 10.1021/jp512994q J. Phys. Chem. B 2015, 119, 5957−5961
Article
The Journal of Physical Chemistry B
N. A. J. The Solution Structure of a Transient Photoreceptor Intermediate: Delta 25 Photoactive Yellow Protein. Structure 2005, 13, 953−962. (12) Nakamura, R.; Hamada, N.; Ichida, H.; Tokunaga, F.; Kanematsu, Y. Coherent Oscillations in Ultrafast Fluorescence of Photoactive Yellow Protein. J. Chem. Phys. 2007, 127, 215102. (13) Mataga, N.; Chosrowjan, H.; Shibata, Y.; Imamoto, Y.; Kataoka, M.; Tokunaga, F. Ultrafast Photoinduced Reaction Dynamics of Photoactive Yellow Protein (PYP): Observation of Coherent Oscillations in the Femtosecond Fluorescence Decay Dynamics. Chem. Phys. Lett. 2002, 352, 220−225. (14) Larsen, D. S.; van Stokkum, I. H. M.; Vengris, M.; van der Horst, M. A.; de Weerd, F. L.; Hellingwerf, K. J.; van Grondelle, R. Incoherent Manipulation of the Photoactive Yellow Protein Photocycle with Dispersed Pump-Dump-Probe Spectroscopy. Biophys. J. 2004, 87, 1858−1872. (15) Nakamura, R.; Hamada, N.; Ichida, H.; Tokunaga, F.; Kanematsu, Y. Ultrafast Dynamics of Photoactive Yellow Protein via the Photoexcitation and Emission Processes. Photochem. Photobiol. 2007, 83, 397−402. (16) Imamoto, Y.; Ito, T.; Kataoka, M.; Tokunaga, F. Reconstitution Photoactive Yellow Protein from Apoprotein and p-Coumaric AcidDerivatives. FEBS Lett. 1995, 374, 157−160. (17) Mihara, K.; Hisatomi, O.; Imamoto, Y.; Kataoka, M.; Tokunaga, F. Functional Expression and Site-Directed Mutagenesis of Photoactive Yellow Protein. J. Biochem. 1997, 121, 876−880. (18) Unno, M.; Kumauchi, M.; Tokunaga, F.; Yamauchi, S. Vibrational Assignment of the 4-Hydroxycinnamyl Chromophore in Photoactive Yellow Protein. J. Phys. Chem. B 2007, 111, 2719−2726. (19) Chosrowjan, H.; Mataga, N.; Shibata, Y.; Imamoto, Y.; Tokunaga, F. Environmental Effects on the Femtosecond-Picosecond Fluorescence Dynamics of Photoactive Yellow Protein: Chromophores in Aqueous Solutions and in Protein Nanospaces Modified by SiteDirected Mutagenesis. J. Phys. Chem. B 1998, 102, 7695−7698. (20) Mataga, N.; Chosrowjan, H.; Taniguchi, S.; Hamada, N.; Tokunaga, F.; Imamoto, Y.; Katoka, M. Ultrafast Photoreactions in Protein Nanospaces as Revealed by fs Fluorescence Dynamics Measurements on Photoactive Yellow Protein and Related Systems. Phys. Chem. Chem. Phys. 2003, 5, 2454−2460. (21) van Wilderen, L. J. G. W.; van der Horst, M. A.; van Stokkum, I. H. M.; Hellingwerf, K. J.; van Grondelle, R. Ultrafast Infrared Spectroscopy Reveals a Key Step for Successful Entry into the Photocycle for Photoactive Yellow Protein. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 15050−15055. (22) Lim, M.; Jackson, T. A.; Anfinrud, P. A. Femtosecond Near-IR Absorbance Study of Photoexcited Myoglobin: Dynamics of Electronic and Thermal Relaxation. J. Phys. Chem. 1996, 100, 12043−12051. (23) Xie, A.; Hoff, W. D.; Kroon, A. R.; Hellingwerf, K. J. Glu46 Donates a Proton to the 4-Hydroxycinnamate Anion Chromophore During the Photocycle of Photoactive Yellow Protein. Biochemistry 1996, 35, 14671−14678.
simultaneously with a 14 ps VC time constant to the surrounding protein through a classical diffusion process. When the IR pump pulse was tuned to 1670 cm−1, the C O stretching modes of the protein backbone were also excited. We observed vibrational energy flow from the CO stretching of Cys69 to the vibrational mode of pCA with a time constant of 4.2 ps.
■
ASSOCIATED CONTENT
* Supporting Information S
Details of the data analysis of the transient absorption peak energies. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jp512994q.
■
AUTHOR INFORMATION
Corresponding Author
*(R.N.) Tel +81-6-6879-7755; Fax +81-6-6879-7878; e-mail r. [email protected]. Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This work was supported by JSPS KAKENHI Grant 24540323 and the Murata Science Foundation. REFERENCES
(1) Leitner, D. M. Energy Flow in Proteins. Annu. Rev. Biochem. 2008, 59, 233−259. (2) Mizutani, Y.; Kitagawa, T. Direct Observation of Cooling of Heme upon Photodissociation of Carbonmonoxy Myoglobin. Science 1997, 278, 443−446. (3) Lian, T.; Locke, B.; Kholodenko, Y.; Hochstrasser, R. M. Energy Flow from Solute to Solvent Probed by Femtosecond IR Spectroscopy: Malachite Green and Heme Protein Solutions. J. Phys. Chem. 1994, 98, 11648−11656. (4) Hamm, P.; Lim, M.; Hochstrasser, R. M. Structure of the Amide I Band of Peptides Measured by Femtosecond Nonlinear-Infrared Spectroscopy. J. Phys. Chem. B 1998, 102, 6123−6139. (5) Kim, Y. S.; Hochstrasser, R. M. Applications of 2D IR Spectroscopy to Peptides, Proteins, and Hydrogen-Bond Dynamics. J. Phys. Chem. B 2009, 113, 8231−8251. (6) Meyer, T. E. Isolation and Characterization of Soluble Cytochromes, Ferredoxins and Other Chromophoric Proteins from the Halophilic Phototrophic Bacterium Ectothiorhodospira-Halophila. Biochim. Biophys. Acta 1985, 806, 175−183. (7) Hoff, W. D.; Düx, P.; Hård, K.; Devreese, B.; Nugteren-Roodzant, I. M.; Crielaard, W.; Boelens, R.; Kaptein, R.; Van Beeumen, J.; Hellingwerf, K. J. Thiol Ester-Linked p-Coumaric Acid as a New Photoactive Prosthetic Group in a Protein with Rhodopsin-Like Photochemistry. Biochemistry 1994, 33, 13959−13962. (8) Baca, M.; Borgstahl, G. E. O.; Boissinot, M.; Burke, P. M.; Williams, D. R.; Slater, K. A.; Getzoff, E. D. Complete ChemicalStructure of Photoactive Yellow Protein - Novel Thioester-Linked 4Hydroxycinnamyl Chromophore and Photocycle Chemist. Biochemistry 1994, 33, 14369−14377. (9) Hellingwerf, K. J.; Hendriks, J.; Gensch, T. Photoactive Yellow Protein, a New Type of Photoreceptor Protein: Will This “Yellow Lab” Bring Us Where We Want to Go? J. Phys. Chem. A 2003, 107, 1082− 1094. (10) Ihee, H.; Rajagopal, S.; Srajer, V.; Pah, R.; Anderson, S.; Schmidt, M.; Schotte, F.; Anfinrud, P. A.; Wulff, M.; Keith Moffat, K. Visualizing Reaction Pathways in Photoactive Yellow Protein from Nanoseconds toSeconds. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 7145−7150. (11) Bernard, C.; Houben, K.; Derix, N. M.; Marks, D.; van der Horst, M. A.; Hellingwerf, K. J.; Boelens, R.; Kaptein, R.; van Nuland, 5961
DOI: 10.1021/jp512994q J. Phys. Chem. B 2015, 119, 5957−5961
