Vibrational Energy Transport in Peptide Helices after Excitation of C

Sep 15, 2009 - Vibrational energy transport in a short 310-helical peptide is studied by time-resolved femtosecond infrared spectroscopy. The C−D vi...
0 downloads 0 Views 650KB Size
Download PDF
J. Phys. Chem. B 2009, 113, 13393–13397

13393

Vibrational Energy Transport in Peptide Helices after Excitation of C-D Modes in Leu-d10 Marco Schade,† Alessandro Moretto,‡ Marco Crisma,‡ Claudio Toniolo,‡ and Peter Hamm*,† Physikalisch-Chemisches Institut, UniVersita¨t Zu¨rich, CH-8057 Zu¨rich, Switzerland, and Institute of Biomolecular Chemistry, PadoVa Unit, CNR, Department of Chemistry, UniVersity of PadoVa, I-35131 PadoVa, Italy ReceiVed: July 6, 2009; ReVised Manuscript ReceiVed: August 13, 2009

Vibrational energy transport in a short 310-helical peptide is studied by time-resolved femtosecond infrared spectroscopy. The C-D vibrations of decadeuterated leucine incorporated in the helical chain are excited, and the subsequent flow of vibrational energy through the helix is monitored by employing CdO probes at various distances from the heat source as local thermometers. The C-D modes are not resonant to the CdO modes, neither directly nor through any Fermi resonance, thereby suppressing resonant energy transfer directly along the CdO oscillators of the peptide backbone. In contrast to our previous work (J. Phys. Chem. B 2008, 112, 9091), we no longer find any substantial difference in the vibrational energy transport efficiency after high- or low-energy excitation. That is, the heat diffusion constant of (2.0 ( 0.5) Å2 ps-1 is the same as that after depositing vibrational energy through the ultrafast internal conversion of a covalently bound chromophore. Introduction Many photochemical reactions in protein environments occur on an ultrafast femtosecond time scale; for instance, the initial isomerization process in bacteriorhodopsin1,2 and rhodopsin,3,4 the photodissociation of CO-myoglobin,5 or excitation quenching in photosynthetic systems.6 A similar situation arises in molecular electronic devices.7 In the simplest picture, one could think of the subsequent dissipation of vibrational energy as heat diffusion. However, temperature gradients of many 100 K are often generated over length scales of a few chemical bonds, and energy is dissipated on a subpicosecond time scale. In this regime, the validity of a heat diffusion equation becomes very questionable. To gain a microscopic understanding of vibrational energy transport phenomena on this sort of length and time scales, several different molecular systems have been investigated, including long-chain hydrocarbon molecules,8 bridged azulene-anthracene compounds,9 small molecules in solution,10 single amino acids,11 phospholipid bilayer liposomes,12 and reverse micelles.13 Regarding energy transport in proteins, a substantial amount of theoretical work has been performed;14-21 however, only very few experimental studies exist, most of which focus on the vibrational energy flow from a heme group into the surrounding solvent.22-25 We have recently introduced a new experimental concept to study vibrational energy transport through short peptide helices.26 Vibrational energy is locally deposited at one end of a peptide 310-helix by ultrafast internal conversion of a UV excited azobenzene moiety. The subsequent energy flow through the helix is detected by using CdO groups in the backbone of the helix as local “thermometers”. The underlying mechanism of this thermometer is discussed in detail in ref 27. Upon heating, vibrational transitions become broader and shift to lower frequency as a result of coupling to vibrationally excited lower frequency modes. To reach site selectivity, certain carbonyl vibrations are singled out from the main amide I band by 13CdO isotope labeling. Placing the isotope label at different positions * Corresponding author. E-mail: [email protected]. † Universita¨t Zu¨rich. ‡ University of Padova.

in the helix, we can measure the local temperature at various distances from the heat source as a function of time (for further details, see refs 26, 28, and 29). We had found in ref 26 that following UV excitation of the azobenzene, the helix is not a particularly good vibrational energy conductor; the heat diffusivity of 2 Å2 ps-1 is significantly slower than values known for bulk materials (10-20 Å2 ps-1 14,15,30,31). In a follow-up paper,28 we studied the dependence of the vibrational energy transport properties on the amount of energy that is initially deposited. That is, we compared the highenergy (UV) excitation result to that after excitation of a single CdO oscillator in the peptide backbone with low-energy (IR) photons and observed that vibrational energy transport was at least four times more efficient in the latter case (heat diffusivity J8 Å2 ps-1). Since this result agreed much better with the findings for bulk materials and also with accompanying MD simulations, we concluded that the IR experiment was the general case of vibrational energy transport in peptides, whereas the UV experiment represented the exception where energy gets trapped because it is deposited in high amounts. The study in ref 28 contains two potential caveats that come along with the use of CdO vibrations as pumping and probing modes at the same time. First, due to the inherent spectral overlap, the probe signals at early times up to about 5 ps are dominated by excitation and relaxation of the directly pumped CdO oscillator and, therefore, cannot be evaluated. As a result, no upper limit for the heat diffusivity could be given. Second, the possibility of resonant energy transfer directly along the CdO oscillators could act as an alternative route of vibrational energy transport. We therefore set out to repeat the low-energy excitation experiment using an alternative heating source for pumping with IR photons. The pumped mode should ideally be nonresonant with the fundamental of the CdO spectator mode used as the local thermometer, as well as with any overtone to also avoid complications with potential Fermi resonances. We chose C-D vibrations of deuterated amino acid side chains, which occur around 2000-2250 cm-1.32-36 The ratio between νC-D and νCdO is about 1.3-1.4 which is close to optimal for avoiding direct

10.1021/jp906363a CCC: $40.75  2009 American Chemical Society Published on Web 09/15/2009

13394

J. Phys. Chem. B, Vol. 113, No. 40, 2009

Schade et al.

as well as Fermi resonances. Furthermore, the frequency lies in an easily accessible spectral range with low background absorption of the solvent chloroform. The C-D vibration is a relatively weak IR absorber, but it has been shown recently by Rubtsov and co-workers11 that the combined intensity of the 10 C-D groups of decadeuterated leucine (Leu-d10) together, all of which can be covered with the bandwidth of our IR pulses, is reasonably large (a total effective extinction coefficient of ≈200 M-1 cm-1). Materials and Methods Laser Setup. The laser pulses for the dual-frequency IR pump/IR probe experiments originate from two white-lightseeded two-stage optical parametric amplifiers (OPA)37 connected to a commercial Ti:sapphire laser/amplifier system. Independently tunable infrared pulses were generated by difference frequency mixing of signal and idler pulses of the OPAs in AgGaS2 crystals (pump: 2 µJ, center frequency 2180 cm-1, bandwidth 200 cm-1; probe: 1.5 µJ, center frequency 1680 cm-1, bandwidth 200 cm-1). The broadband pump pulses were focused on the sample and excited all 10 C-D vibrations of the leucine simultaneously. The probe pulses were focused into the sample in spatial overlap with the pump pulses; a reference beam, split off beforehand, was focused about 1 mm separated into a nonexcited volume of the sample. Both probe and reference beams were frequency-dispersed in a spectrometer and imaged on a double-array HgCdTe detector with a spectral resolution of ≈4 nm. The pump pulse was delayed with respect to the probe pulse by an optical delay stage. The overall instrumental response time was 300 fs. All beams were polarized parallel. Peptide Helix. The two terminally protected, heptameric peptides used in this study are Z-Aib-L-Leu**-Aib-L-Ala*(Aib)3-OMe (compound 1) and Z-Aib-L-Leu**-(Aib)3-L-Ala*Aib-OMe (compound 2), where one asterisk (*) indicates a 13 CdO label, two asterisks (**) stand for decadeuteration, Z is benzyloxycarbonyl, Aib is R-aminoisobutyric acid, and OMe is methoxy (Figure 1a). As in our previous studies,26,28,29 the isotope labels are placed by incorporating 13C-L-Ala, since this amino acid is readily available commercially and does not destabilize the 310-helix significantly (see also the 3D-Structural Characterization section). This particular amino acid sequence, which is different from our previous one,26,28,29 was chosen according to a few stringent considerations. Adding an additional Leu-d10 to the original octameric sequence would have led to a nonameric peptide, which is close to the limit for a sufficient solubility (particularly in CHCl3 solution). Furthermore, the coexistence of 310- and R-helices is very likely for Aib-rich peptides of that size.38-40 To maintain the pure 310-helical structure, therefore, we decided to shorten the sequence. Finally, an N-terminal -L-Leu-Aib- sequence should be avoided because, in that case, this dipeptide could fold in either a nonhelical typeII β-turn or a helical type-III β-turn.38,40 On these bases, we decided to use the 7-mers with the sequences mentioned above. Peptide synthesis was performed in solution by activating the carboxyl function with 1-(3-dimethylamino)propyl-3-ethylcarbodiimide and 7-aza-1-hydroxy-1,2,3-benzotriazole.41 Details of the syntheses and characterizations are reported in the Supporting Information. The peptides were dissolved in chloroform at a concentration of ≈10 mM. The sample was kept between 2-mm-thick CaF2 windows separated by a 300 µm spacer. All measurements were carried out at room temperature.

Figure 1. (a) Chemical structure of the heptapeptide, here exemplified for 1. The labels indicate the positions of the different CdO vibrators, which are used as local thermometers, relative to the position of Leud10. In 1, the 13C-labeled Ala is situated at site #3 (as shown here); in 2, the label is placed at site #5. (b) X-ray diffraction structures of the two independent peptide molecules (A and B) of 1 in the asymmetric unit. H and D atoms have been omitted for clarity. Intramolecular H-bonds are indicated by dashed lines.

Experimental Results 3D-Structural Characterization. X-ray diffraction, FTIR absorption, 1H NMR, and CD techniques were used to examine the preferred conformations of the NR-acylated heptapeptide esters 1 and 2. In our hands, only 1 gave a single crystal suitable for an X-ray diffraction analysis. Interestingly, there are two independent peptide molecules in the asymmetric unit of 1 (Figure 1b and Supporting Information). Both are regular 310helices, each stabilized by five intramolecular i r i + 3 CdO · · · H-N H-bonds. The main difference between them is found in their helical screw sense: one molecule (A) is righthanded, whereas the other (B) is left-handed (in other words, the two molecules are diastereomers). This is a rather uncommon but not unique observation in peptide crystallography (see also refs 26 and 42). What is unique, to the best of our knowledge, is that this peptide sequence is characterized by the occurrence of as many as two chiral L-residues, not by one, as found in all previously observed diastereomeric pairs. The FTIR absorption spectra in the conformationally informative N-H stretching region of peptides 1 and 2 in the structure-supporting solvent chloroform (Figure S1 of the Supporting Information) are similar, showing a weak band at about 3425 cm-1 (free NH groups) and a very strong band in the 3320-3330 cm-1 region (hydrogen-bonded NH groups).43 We assign the observed hydrogen bonding to the intramolecular type because the spectra do not significantly change upon variation of the peptide concentration between 1.0 and 0.1 mM (not shown). The ratio of the integrated intensity of the band of hydrogen-bonded NH groups to that of free NH groups is remarkably high for both peptides 1 and 2, suggesting the onset of rather stable helical structures. In addition, for 1, this ratio is close to that previously reported for the 310-helical -(Aib)7-

Vibrational Energy Transport in Peptide Helices

J. Phys. Chem. B, Vol. 113, No. 40, 2009 13395

Figure 2. Absorption spectrum of 1 in the spectral region of the CdO and the C-D stretch vibrations. The C-D bands are scaled up by a factor of 10 for better visibility (blue).

homopeptide43 and is slightly higher than that of 2. The latter finding is indicative of a marginal destabilization induced in the helix by the L-Ala residue when inserted in the penultimate position of the peptide chain, as compared to a central position. The results of the two 1H NMR titrations in deuterochloroform solution upon addition of the hydrogen bonding acceptor dimethylsulfoxide (DMSO) (Figures 2S and 3S of the Supporting Information) clearly show that two NH protons for each heptapeptide are exposed to the solvent, whereas the remaining five NH protons are solvent-protected by hydrogen bonding. One of these protons (that at the highest field) is easily assigned to the N-terminal Aib proton by virtue of its urethane character43 and multiplicity (singlet). The second solvent-exposed NH proton is attributed to the Leu-d10 at position 2 in the peptide chain on the basis of a number of conformational studies published on Aib-rich, medium-sized, peptides.26,38,39,43 These overall 1H NMR properties are typical for peptides folded in a 310-helical structure.43 The far-UV CD spectra of the two heptapeptides in 2,2,2trifluoroethanol, also a structure-supporting solvent (Figure 4S of the Supporting Information), exhibit two negative Cotton effects (222-230 nm, weak; and 202-203 nm, very strong), and the ratios of their intensities are in the range expected for peptides largely adopting the 310-helical structure.44 From this extensive conformational analysis, we can safely conclude that both heptapeptides 1 and 2 are overwhelmingly 310-helical, although not to the same extent, under structurally favorable experimental conditions. Steady-State Spectra. Figure 2 shows the stationary absorption spectrum of the peptide helix in the spectral region of the CdO and the C-D stretch vibrations (exemplified here for 1; the spectrum of 2 is almost identical). In the region around 2200 cm-1, five C-D bands belonging to Leu-d10 appear. According to ref 11, they can be assigned as follows: 2065 cm-1, CδD3 (ss); 2101 cm-1, CβD2 (ss); 2126 cm-1, CγD; 2175 cm-1, CβD2 (as), and 2221 cm-1, CδD3 (as). The CdO bands in the 1680 cm-1 region can be assigned as described in refs 26, and 28. The intense band at 1665 cm-1 is the “main band” consisting of all equivalent CdO groups in the helix (five in total). The isotope-labeled 13CdO band (CdO #3 in 1, CdO #5 in 2) appears red-shifted with respect to the main band at about 1620 cm-1. The bands at 1715 and 1734 cm-1 belong to the two chemically different CdO groups; namely, the urethane moiety connected to the N-protecting group (CdO #-2) and the ester group of the C-terminal Aib residue (CdO #6). Time-Resolved Spectra. The time-resolved IR pump/IR probe spectra from the CdO stretching region are displayed in Figure 3 (a linear baseline determined from the edges of the spectral window was subtracted from all spectra). Upon excitation of the Leu-d10 C-D vibrations, all CdO modes exhibit

Figure 3. Transient spectra in the CdO stretching region for 1 (left column) and 2 (right column) recorded at various delay times (green, 1.00 ps; red, 1.75 ps; blue, 2.75 ps; black, 10 ps; orange, 50 ps) after pumping of the leucine C-D vibrations around 2200 cm-1. The labels refer to the positions of the CdO groups relative to the heater (see also Figure 1a). The black dashed line is a stationary temperatureinduced difference spectrum corresponding to a temperature jump of 7.5 mK. For comparison, the stationary absorption spectra of 1 (left column) and 2 (right column) are plotted above.

red shifts, which are due to anharmonic coupling of the spectator modes to thermally excited lower frequency modes.27 Since the resulting (negative) bleach signal is, in general, narrower than the anharmonically shifted (positive) hot band, we use the former as a measure of the amount of vibrational energy in the vicinity of the probe. As expected, the main band at 1665 cm-1 shows the strongest response. This is because it averages over the five nonlabeled CdO oscillators in the helix, among which are the two direct neighbors of the vibrationally excited Leu-d10; that is, sites #-1 and #1. Comparing the amplitudes of the bleach signals at 1620 cm-1, belonging to the isotope labels at site #3 (compound 1, Figure 3, left) and site #5 (compound 2, Figure 3, right), it immediately becomes clear that more vibrational energy reaches the closer site #3. Interestingly, the response of site #-2 is more than twice as strong as that of site #3, although its distance to the heater is almost the same (5 vs 7 chemical bonds; see Figure 1a). This is due to a boundary effect, which pools vibrational energy at the terminal site of the chain (see discussion below). The same effect, albeit less pronounced, can be observed for the C-terminal site #6. The spectra recorded at late delay times (Figure 3, orange lines) resemble a stationary temperature difference spectrum (Figure 3, black dashed lines). The latter has been measured for a 10 K temperature increase in a FTIR spectrometer and is down-scaled in Figure 3 to match the 50 ps time-resolved result. From the scaling factor, we estimate that the bulk solvent has been heated up by approximately 7.5 mK. As a secondary effect, this temperature rise leads to an overall weakening of the hydrogen bonds of the helix, resulting in a blue shift of the CdO bands.45 The kinetics of the main band bleach is shown in Figure 4 as a cut through the transient spectra of 1 at 1673 cm-1. A biexponential fit reveals a rising time of 1.1 ps and a decay time of 7.6 ps. The bleach of the main band is dominated by the probes next to the leucine heater; that is, sites #-1 and #1. Assuming that vibrational energy propagation from the Leud10 side chains to the peptide backbone is not the rate-limiting step, the rising time of the main band is expected to essentially reflect the vibrational lifetime of the C-D modes in Leu-d10. The latter was measured to be T1 ) 1.3 ps,11 in good agreement with our result. The ≈7 ps decay component, on the other hand, is known as a typical cooling time in chloroform46 and is consistent with all our previous studies.26,28,29

13396

J. Phys. Chem. B, Vol. 113, No. 40, 2009

Figure 4. Time dependence of the bleach intensity of the main band (recorded for 1) with opposite sign as in Figure 3. The experimental data can be fitted with a biexponential function with a rising time of 1.1 ps and a decay time of 7.6 ps.

Figure 5. (a) Scheme of the rate model used for global fitting of the experimental data. The Leu-d10 heater is directly coupled to its next neighbors. The corresponding rate constant is given by the inverse lifetime of the C-D modes; that is, kCD)(1.1 ps)-1. Cooling of all peptide units into the solvent occurs with a rate constant of ks ) (7.6 ps)-1; the propagation rate in the helix is kp ≈ (2 ps)-1. (b) Time dependence of the bleach intensity of sites #-2, #3 (both recorded at 1), and #5 (recorded at 2) with opposite signs as in Figure 3. The solid lines represent global fits according to the rate model shown on top.

The kinetics of sites #-2, #3, and #5 are summarized in Figure 5b. One can nicely see the sequential appearance of the bleach maxima, showing that vibrational energy propagates through the molecule. The farther a reporter unit is away from the heater, the later and the less vibrational energy arrives. To model the experimental data, we constructed a rate equation scheme illustrated in Figure 5a. Vibrational energy is initially deposited in the C-D modes of the Leu-d10 heater, which decay with a characteristic lifetime T1. As a result, vibrational energy is transferred to the next neighbors (i.e., sites #-1 and #1) with a rate constant kCD ) T1-1. Each peptide unit then can exchange energy with its nearest neighbors with a propagation rate kp. Furthermore, all peptide units lose energy into the solvent at rate ks. The rate constants kCD and ks were taken from the main band fit: (1.1 ps)-1 and (7.6 ps)-1, respectively. A global fit of the time traces of sites #-2, #3, and #5 then yields a propagation rate constant kp ) (2.0 ( 0.5)-1 ps-1 (see Figure 5b, solid lines). Discussion The experimental data can be reasonably described by a diffusive model. One can think of the rate equation model of Figure 5a as a discretized version of a diffusion equation, in

Schade et al. which the diffusion constant, D, is related to the propagation rate, kp, through D ) kp∆x2, with the helical translation per residue for a 310-helix of ∆x ≈ 2 Å.38,47 Hence, we obtain a heat diffusion constant of (2.0 ( 0.5) Å2 ps-1, which is in quantitative agreement with our previous UV pump study (2 Å2 ps-1).26 Thus, within signal-to-noise, no excitation energy dependence of the heat diffusion constant is observed. The value we obtain is much slower than that after IR pumping of backbone CdO groups (J8 Å2 ps-1).28 This shows that from the three possible explanations we offered in ref 28 for the enhanced propagation rate after IR pumping, it is the resonant energy transfer directly along the CdO oscillators that bypasses vibrational energy flow along the peptide backbone. We obviously had underestimated the significance of this process in ref 28. In essence, there is no fundamental difference between the transport properties after high- and low-energy excitation when suppressing the additional route of resonant energy transfer, and the linearity of vibrational energy diffusion is maintained, despite the fact that temperature gradients of many hundreds of Kelvin are generated over length scales of only a few chemical bonds in the UV pump experiment. This interpretation is also supported by our recent theoretical study on this subject,48 showing that the energy transport properties of molecular chains are largely insensitive to the amount of initially deposited energy. Furthermore, other experimental studies11 also report propagation effects after IR excitation that happen on a comparably slow time scale. This leaves us with the question why the apparent heat diffusion constant we observe (≈2.0 Å2 ps-1) is much slower than values known from bulk materials (10-20 Å2 ps-1 14,15,30,31). In a recent simulation study,48 we proposed that the rate-limiting step in vibrational energy transport on these short length scales might be intrasite vibrational energy redistribution (IVR) rather than the actual transport along the chain. In simple words, vibrational energy does not completely thermalize within individual sites on the picosecond time scale of transport. Due to the large number of modes involved, IVR has probabilistic character, and vibrational energy transport will be diffusivelike, in the sense that it can be fitted by a rate equation system. However, since intrasite IVR is chain-length-independent, at some point, transport along the chain will become rate-limiting, thereby approaching the limit of an infinitely long peptide. Hence, the apparent diffusion constant we observe in this type of experiment does not necessarily reflect the value observed in macroscopic experiments. Further, it should be borne in mind that the geometry of the helix more resembles a cylinder rather than a wire. Due to the special boundary conditions imposed in this geometry, energy transport properties can additionally deviate from the simple 1D case, since two length and time scales get involved (radial and longitudinal diffusion). Nevertheless, since energy transport in our case is dominated by IVR, considering the effects of a cylindrical geometry would overstress the heat diffusion equation. It is interesting to note that the rate equation model can nicely reproduce the effect of energy pooling at the N-terminal site #-2. The response of site #-2, measured relative to that of site #3, is much larger than what would be expected just from their different distances (5 vs 7 chemical bonds; see Figure 1a). From the rate equation model, one expects the response of site #-2 to be about 50% larger than the response of site #2, although both are situated at the same distance from the heater. This is the result of a boundary effect for the diffusion along a finite 1D chain, which provides additional strong evidence that we, indeed, observe vibrational energy transport along the

Vibrational Energy Transport in Peptide Helices peptide backbone, and not through the solvent. Actually, the bleach of site #-2 should be even more distinct than found in the experiment. The fact that the simulation slightly overestimates its intensity could be due to the N-protecting group (Zmoiety); hence, site #-2 is not really the boundary of the chain. Conclusion In the present paper, we have shown that a Leu-d10 residue provides a feasible local heater initiating vibrational energy transport through the peptide backbone. Decadeuterated leucine is commercially available and can be easily incorporated into a peptide sequence by standard methods of peptide synthesis. Although the C-D vibrations are relatively weak IR absorbers, the total extinction coefficient of all 10 leucine C-D vibrations together is large enough to deposit a sufficient amount of vibrational energy, whose propagation through the peptide helix then can be followed with a reasonable signal-to-nose ratio. Vibrational energy propagation after IR excitation of the C-D modes can be described as a diffusion-like process at room temperature. The effect of terminal energy pooling provides further evidence that energy propagates along the molecular chain, and not through the solvent. We find that vibrational energy is transported relatively slowly, with an apparent heat diffusion constant of ≈2 Å2 ps-1, in agreement with our previous UV pump experiment.26 Contrary to our previous conclusion, we now attribute the 4-fold faster energy transport we observed in ref 28 to resonant energy transfer directly along the CdO oscillators. In the present case, this pathway is strongly suppressed because the C-D vibrations are nonresonant to the amide I vibrations. Since the amount of initially deposited energy does not significantly affect the subsequent equilibration process, energy transport properties can be studied equally well by UV pumping. From a practical point of view, this is advantageous because much more energy per molecule can be deposited, revealing much larger signals that can be observed for longer distances. This opens the possibility to extend vibrational energy transport studies to proteins to experimentally address the question whether specific transport pathways indeed exist, because they are consistently predicted from MD simulations.14,15,19,49-55 Acknowledgment. We thank Gerhard Stock for numerous very instructive discussions on the topic. This work has been supported by the Swiss National Science Foundation under Grant 200021-124463/1. Supporting Information Available: Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Mathies, R.; Cruz, C. H. B.; Pollard, W. T.; Shank, C. V. Science 1988, 240, 777. (2) Dobler, J.; Zinth, W.; Kaiser, W.; Oesterhelt, D. Chem. Phys. Lett. 1988, 144, 215. (3) Wang, Q.; Schoenlein, R. W.; Peteanu, L. A.; Mathies, R. A.; Shank, C. V. Science 1994, 266, 422. (4) Kukura, P.; McCamant, D. W.; Yoon, S.; Wandschneider, D. B.; Mathies, R. A. Science 2005, 310, 1006. (5) Champion, P. M. Science 2005, 310, 980. (6) Cerullo, G.; Polli, D.; Lanzani, G.; Silvestri, S. D.; Hashimoto, H.; Cogdell, R. J. Science 2002, 298, 2395. (7) Chang, C. W.; Okawa, D.; Majumdar, A.; Zettl, A. Science 2006, 314, 1121. (8) Wang, Z.; Carter, J. A.; Lagutchev, A.; Koh, Y. K.; Seong, N. H.; Cahill, D. G.; Dlott, D. D. Science 2007, 317, 787. (9) Schwarzer, D.; Kutne, P.; Schro¨der, C.; Troe, J. J. Chem. Phys. 2004, 121, 1754.

J. Phys. Chem. B, Vol. 113, No. 40, 2009 13397 (10) Wang, Z.; Pakoulev, A.; Dlott, D. D. Science 2002, 296, 2201. (11) Naraharisetty, S. R. G.; Kasyanenko, V. M.; Zimmermann, J.; Thielges, M. C.; Romesberg, F. E.; Rubtsov, I. V. J. Phys. Chem. B 2009, 113, 4940. (12) Kuciauskas, D.; Wohl, C. J.; Pouy, M.; Nasai, A.; Gulbinas, V. J. Phys. Chem. B 2004, 108, 15376. (13) Dea`k, J. C.; Pang, Y.; Sechler, T. D.; Wang, Z.; Dlott, D. D. J. Phys. Chem. B 2004, 306, 473. (14) Yu, X.; Leitner, D. M. J. Phys. Chem. B 2003, 107, 1698. (15) Yu, X.; Leitner, D. M. J. Chem. Phys. 2005, 122, 054902. (16) Henry, E. R.; Eaton, W. A.; Hochstrasser, R. M. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 8982. (17) Tesch, M.; Schulten, K. Chem. Phys. Lett. 1990, 169, 97. (18) Okazaki, I.; Hara, Y.; Nagaoka, M. Chem. Phys. Lett. 2001, 337, 151. (19) Sagnella, D. E.; Straub, J. E. J. Phys. Chem. B 2001, 105, 7057. (20) Nguyen, P. H.; Gorbunov, R. D.; Stock, G. Biophys. J. 2006, 91, 1224. (21) Fujisaki, H.; Straub, J. E. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 6726. (22) Miller, R. J. D. Annu. ReV. Phys. Chem. 1991, 42, 581. (23) Lian, T.; Locke, B.; Kholodenko, Y.; Hochstrasser, R. M. J. Phys. Chem. 1994, 98, 11648. (24) Li, P.; Champion, P. M. Biophys. J. 1994, 66, 430. (25) Mizutani, Y.; Kitagawa, T. Science 1997, 278, 443. (26) Botan, V.; Backus, E. H. G.; Pfister, R.; Moretto, A.; Crisma, M.; Toniolo, C.; Nguyen, P. H.; Stock, G.; Hamm, P. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 12749. (27) Hamm, P.; Ohline, S. M.; Zinth, W. J. Chem. Phys. 1997, 106, 519. (28) Backus, E. H. G.; Nguyen, P. H.; Botan, V.; Pfister, R.; Moretto, A.; Crisma, M.; Toniolo, C.; Stock, G.; Hamm, P. J. Phys. Chem. B 2008, 112, 9091. (29) Backus, E. H. G.; Nguyen, P. H.; Botan, V.; Pfister, R.; Moretto, A.; Crisma, M.; Toniolo, C.; Zerbe, O.; Stock, G.; Hamm, P. J. Phys. Chem. B 2008, 112, 15487. (30) Lide, D. R. CRC Handbook of Chemistry and Physics, 75th ed.; CRC Press: Boca Raton, FL, 1994. (31) Touloukian, Y. P.; Powelland, R. W.; Ho, C. Y.; Nicolaou, M. C. Thermal Physical Properties of Matter; Plenum: New York, 1978. (32) Chin, J. K.; Jimenez, R.; Romesberg, F. E. J. Am. Chem. Soc. 2001, 123, 2426. (33) Sagle, L. B.; Zimmermann, J.; Dawson, P. E.; Romesberg, F. E. J. Am. Chem. Soc. 2004, 126, 3384. (34) Gundogdu, K.; Nydegger, M. W.; Bandaria, J. N.; Hill, S. E.; Cheatum, C. M. J. Chem. Phys. 2006, 125, 174503. (35) Kumar, K.; Sinks, L. E.; Wang, J. P.; Kim, Y. S.; Hochstrasser, R. M. Chem. Phys. Lett. 2006, 432, 122. (36) Zimmermann, J.; Gundogdu, K.; Cremeens, M. E.; Bandaria, J. N.; Hwang, G. T.; Thielges, M. C.; Cheatum, C. M.; Romesberg, F. E. J. Phys. Chem. B 2009, 113, 7991. (37) Hamm, P.; Kaindl, R. A.; Stenger, J. J. Opt. Lett. 2000, 25, 1798. (38) Toniolo, C.; Crisma, M.; Formaggio, F.; Peggion, C. Biopolymers 2001, 60, 396. (39) Karle, I. L.; Balaram, P. Biochemistry 1990, 29, 6747. (40) Prasad, B. V. V.; Balaram, P. CRC Crit. ReV. Biochem. 1984, 16, 307. (41) Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397. (42) Valle, G.; Crisma, M.; Toniolo, C.; Beisswenger, R.; Rieker, A.; Jung, G. J. Am. Chem. Soc. 1989, 111, 6828. (43) Toniolo, C.; Bonora, G. M.; Barone, V.; Bavoso, A.; Benedetti, E.; Di Blasio, B.; Grimaldi, P.; Lelj, F.; Pavone, V.; Pedone, C. Macromolecules 1985, 18, 895. (44) Toniolo, C.; Polese, A.; Formaggio, F.; Crisma, M.; Kamphuis, J. J. Am. Chem. Soc. 1996, 118, 2744. (45) Pimentel, G. C.; McClellan, A. The Hydrogen Bond; Freeman: San Francisco, 1960. (46) Dahinten, T.; Baier, J.; Seilmeier, A. Chem. Phys. Lett. 1998, 232, 239. (47) Toniolo, C.; Benedetti, E. Trends Biochem. Sci. 1991, 16, 350. (48) Schade, M.; Hamm, P. J. Chem. Phys. 2009, 131, 044511. (49) Ota, N.; Agard, D. A. J. Mol. Biol. 2005, 351, 345. (50) Sharp, K.; Skinner, J. J. Proteins 2006, 65, 347. (51) Leitner, D. M. AdV. Chem. Phys. 2005, 130B, 205. (52) Leitner, D. M. Annu. ReV. Phys. Chem. 2008, 59, 233. (53) Ishikura, T.; Yamato, T. Chem. Phys. Lett. 2006, 432, 533. (54) Gao, Y.; Koyama, M.; El-Mashtoly, S. F.; Hayashi, T.; Harada, K.; Mizutani, Y.; Kitagawa, T. Chem. Phys. Lett. 2006, 429, 239. (55) Moritsugu, K.; Miyashita, O.; Kidera, A. Phys. ReV. Lett. 2000, 85, 3970.

JP906363A