J. Phys. Chem. B 2010, 114, 15151–15157
15151
Hydration Effects on Energy Relaxation of Ferric Cytochrome C Films after Soret-Band Photoexcitation Shuji Ye*,† and Andrea Markelz*,‡ Hefei National Laboratory for Physical Sciences at Microscale, UniVersity of Science and Technology of China, Hefei, Anhui, People’s Republic of China 230026, and Department of Physics, UniVersity at Buffalo, SUNY, 239 Fronczak Hall, Buffalo, New York 14260-1500, United States ReceiVed: May 10, 2010; ReVised Manuscript ReceiVed: September 27, 2010
Protein hydration plays a critical role in protein dynamics and biological processes. Pump-probe transmission measurement has been applied to investigate the hydration effects on the energy relaxation of a heme protein ferric Cytochrome c (Cyt c) film after soret-band photoexcitation. Transient dynamics study indicates that the energy internal conversion time of ∼300 fs is independent of hydration. The vibrationally excited electronic ground-state recovery rates show two transitions at the hydration level of h ) 12.4-16.5% and 21.7-23.5%. The first transition occurs at the hydration level for the onset of an increasing ferric Cyt c flexibility while the second transition occurs at the saturated hydration level. The hydration dependence of steady-state electronic absorption spectrum results shows that the Q-band peak is nearly constant in center wavelength, but the line width surprisingly narrows with increasing hydration. For the ∼695 nm absorbance associated with the MET80Fe bond, the intensity increases with increasing hydration and slightly blue shifts. The 695 nm peak grows rapidly at h ) 12.4% and then plateaus at h ) 21.7%. This research shows that ∼695 nm absorbance and ground-state recovery rates are sensitive to the hydration of the protein. This study will aid in understanding how hydration modulates the activity of the protein dynamics at a local level. 1. Introduction It has been universally recognized that protein hydration is essential for the structure, stability, dynamics, and function of proteins and other biological macromolecules.1-20 The specific heat, the thermal stability, and the glasslike transition of proteins correlate strongly with their hydration level. For many proteins, function requires a critical hydration level.1,2 The origin of this requirement has been considered by a number of techniques, such as calorimetry,21 infrared spectroscopy,22 osmotic pressure analysis,23 NMR techniques,24-26 dielectric spectroscopy,27,28 time-resolved fluorescence studies,17 neutron diffraction,29 and so forth. These studies have shown a correlation between hydration and static protein three-dimensional structure. However, most of these studies on hydration dependence have focused on global motion changes of the proteins, not on the effects at a local level. It has become increasingly clear that change in static three-dimensional structure alone is insufficient to explain the molecular mechanism of the hydration effects on function. Recently, Zhong et al. investigated protein hydration dynamics by integrating ultrafast fluorescence spectroscopy with the site-directed mutagenesis method. Their studies showed a strong correlation between local solvation dynamics and peptide/ protein conformation transitions and the critical role of hydration water in the structural integrity of the peptide/protein.30-33 Cytochrome c (Cyt c), a model heme protein, plays a central role in the oxidation-reduction reactions of living systems.34 It transfers electrons from the dehydrogenases to terminal acceptors, a process that drives oxidative phosphorylation and * To whom correspondence should be addressed. (S.Y.) E-mail:
[email protected]. Fax: 86-551-3603462. Tel: 86-551-3603462. (A.M.) E-mail:
[email protected]. Fax: (716) 645-2507. Tel: (716) 645-2017, ext. 124. † University of Science and Technology of China. ‡ University at Buffalo, SUNY.
ATP synthesis in the mitochondria of the cells.35 It has been widely used to study the photoactivation of heme proteins and their energy relaxation,35-41 as well as refolding and unfolding.42-48 The hydration of Cyt c under static conditions has been studied by dielectric spectroscopy,49,50 and terahertz absorption spectroscopy.51 The dielectric studies based on global motion changes suggest that there are two classes of protein-bound water molecules (rotationally hindered and unhindered). The bound water acts as a plasticizer, resulting in an increase in free volume and greater macroscopic mobility. In addition, water molecules bound to the protein may therefore increase the local dielectric constant near charged residues, thus reducing the net electrostatic force between charges and enabling greater flexibility.49,52 For Cyt c, the hydration level of tightly bound water is 7.9 ((0.1)% g water/g protein and the level for the onset of an increasing protein flexibility is 14.5 ((0.1)%.49 The flexibility saturates at 26.1 ((5.8)% hydration, which we will refer to as the saturation hydration.50 The photoactivation of Cyt c solution and its energy relaxation has been well studied.33,35-41 However, how the hydration affects the energy relaxation has not been previously addressed. A detailed characterization of the energy relaxation under different hydration level is a fundamental step in the understanding of the hydration effects on protein dynamics and functions at a local level. Here we study the hydration dependence of energy relaxation in ferric Cyt c films by using pump-probe transmission measurements with excitation at the Soret band (400 nm) and probing at 400 and 800 nm. In addition, we also investigate the hydration dependence of steady-state absorption. This study will aid in understanding how hydration modulates the activity of the protein dynamics at a local level.
10.1021/jp104217j 2010 American Chemical Society Published on Web 10/28/2010
15152
J. Phys. Chem. B, Vol. 114, No. 46, 2010
Ye and Markelz
Figure 1. Two-color pump-probe transmission experimental setup. BS, beam splitter; M1, M2, M3, M4, M5, Mirror; L, lense; P, polarizer.
2. Materials and Methods 2.1. Sample Preparation. Horse heart ferric Cyt c (Sigma Chemical Co.) was used as purchased without further purification. The films were prepared by pipetting 1.0 mM ferric Cyt c/Trisma-buffer solution (pH 7.0) onto infrasil quartz substrates. The films were formed by controlled drying.51 Two representative films are discussed here. A thin film with an optical density (O.D.) of 1.5 ((0.1) at 400 nm at hydration level of 8.4% g water/g protein, was used for the single-color pump-probe measurements and steady state absorption measurements in Soret and Q-band region. The thick film with an O.D. of 0.5 ((0.1) at 800 nm at hydration level of 8.4% g water/g protein was used for two-color pump-probe measurements and steady state absorption measurements from 570 to 850 nm. The O.D. at 555 nm is 0.19((0.02) and 1.4 ((0.1) for the thin and thick film, respectively. 2.2. Experimental Methods. All of the experiments were performed at room temperature (T ) 23 °C). Figure 1 shows the two-color pump-probe transient absorption measurement setup. Femtosecond pulses were generated by a mode-locked Ti: Sapphire laser with regenerative amplifier system (Spitfire, Spectra Physics, CA) that yields ∼140 fs pulses at 800 nm with an energy of 850 µJ at a repetition rate of 1 kHz. This system has been applied to study the ultrafast carrier dynamics.53 The pulses passed through a telescope with convex and concave mirrors to produce 2 mm spot size. In two-color pump-probe experiments, the pulses were split into two paths using a beam splitter (BS). One beam passed through BBO/filter system in which second harmonic light at 400 nm was obtained by a 0.1 mm β-barium-borate (BBO) crystal and the fundamental light (800 nm) was rejected by a glass filter (BG 39, Schott). The 400 nm pulses with pulse widths of about 150 fs were used as the pump pulses and chopped at 500 Hz. Another beam (800 nm) was used as probe pulses and passed through a delay stage. The pump and probe light both were focused on the samples with a spot size of 1.1 mm for pump and 0.9 mm for probe light. The transmission change was detected by amplified silicon PIN photodiode (PDA 520, Thorlabs) as a function of time delay between the pump and probe beams. The single-color experiment setup is similar to Figure 1 except that BBO/filter system was moved to the space between M1 and BS in Figure 1. The steady-state absorption spectrum was recorded by HR4000CG-UV-NIR broadband spectrometer (Ocean Optics). The continuum light was generated by Fiber-Lite PL-900 (Dolan-Jenner). In the all experiments, the film samples were mounted in a hydration cell and hydrated for more than 1 h before running
the experiments. Relative humidity was controlled by a LI-610 portable Dew Point Generator (LI-COR). The mass percent water content of ferric Cyt c film (g water/g protein) was determined by using the isotherm equation reported by Gascoyne and Pethig54
h)
Vmax (1 - bx)[1 + (a - b)x]
(1)
where Vm is the monolayer hydration value, a is the activity of gas absorbed in the first layer, b is the activity of gas absorbed in subsequent layers, and x is the relative humidity. Here, Vm ) 8.27, a ) 13.4, and b ) 0.88 for Cyt c. The primary source of uncertainty was due to drift in the relative humidity within the sample cell arising from fluctuations of the ambient temperature of (1 °C during a complete hydration measurement. 3. Results and Discussion 3.1. Hydration Dependence of Steady-State Absorption Spectra of Ferric Cyt c. The steady-state electronic absorption spectrum of ferric Cyt c is dominated by the near-UV and visible bands, usually called the Soret (or B) band, and Q bands, corresponding to the in-plane π f π* transitions of the porphyrin ring.55 The Soret band occurs at 410 nm and Q-band shows two peaks at 530 and 565 nm.55 The spectrum also shows a weaker absorption at 695 nm, which arises from the promotion of a porphyrin π orbital to the dz2 orbital of the low-spin iron and is associated with the ligation of the heme iron by methionine-80 sulfur (MET80).55-61 Here we refer to the peak at 695 nm as MET80-Fe absorption peak. The measured hydration dependence of the steady state electronic absorption spectrum is presented for Soret and Q bands in Figure 2 and in near-infrared range from 570 to 850 nm in Figure 3. Within the precision of our measurements, there is no apparent hydration dependence of the central peak position and intensity of the Soret and Q bands. However, as seen in Figure 3, while the peak position of the Q-band does not change, the apparent line width decreases with increasing hydration whereas the intensity of the MET80-Fe absorption increases. This perhaps indicates that the structure of MET80-Fe group changes. To describe the hydration dependence observed, we fit the data between 630 and 770 nm using two Gaussians: one centered in the 530 nm range to fit the broad absorption from 630 to 770 nm, which may arise from Q-band, C-T bands attributed to π f Fe(dyz) or iron d f d transitions, and the second centered at ∼695 nm which is associated with the
Energy Relaxation of Ferric Cytochrome C Films
J. Phys. Chem. B, Vol. 114, No. 46, 2010 15153
Figure 5. Hydration dependence of the amplitude of the absorption peak centered at ∼695 nm.
Figure 2. Hydration (g water/g protein) dependence of absorption in Soret and Q-band region: (a) h ) 7.6%; (b) h ) 9.5%; (c) h ) 12.4%; (d) h ) 16.5%; and (e) h ) 23.5%.
Figure 3. Hydration (g water/g protein) dependence of absorption in near-infrared range from 570 to 850 nm: (a) h ) 7.6%; (b) h ) 12.4%; (c) h ) 16.5%; (d) h ) 23.5%; and (e) h ) 27.2%.
MET80-Fe bond. The fits are shown in Figure 4. The hydration dependence of the total area of the fitting peak centered at∼695 nm is shown in Figure 5. To ensure the apparent hydration trend in Figure 5 is not an artifact of the removal of the background from the broad peak centered at ∼530 nm, we varied the peak center of the broad peak from 500-560 nm. However, the fitting is not noticeably affected by the variation in background fitting. With hydration increasing, the Q-band peak is nearly constant in center wavelength, but the line width surprisingly narrows. For the ∼695 nm absorbance, the total area increases with increasing hydration and slightly blue shifts. The 695 nm peak grows rapidly at h ) 12.4% and then plateaus at h ) 21.7%. Two apparent transitions are observed at the hydration level of 12.4-16.5 and 21.7-23.5%. The first transition occurs at the hydration level at which protein flexibility begins to increase while the second one occurs at the saturated hydration level.49,50,52 The results suggested that the absorption area of MET80-Fe band is strongly affected by ferric Cyt c flexibility. Earlier reports indicate that the peak at 695 nm is very sensitive to the structure and thermodynamic stability of the protein.55-63 Although high-resolution X-ray studies on oxidized64 and reduced65 tuna cytochrome c have revealed that the heme group geometry is distorted from ideal D4h symmetry, the heme group in Cyt c can approximately be assumed to have D4h symmetry.66,67
Figure 4. Fitting results for the absorption between 630 and 770 nm in Figure 3: (a) h ) 7.6%; (b) h ) 12.4%; (c) h ) 16.5%; (d) h ) 21.7%; (e) h ) 23.5%; and (f) h ) 27.2%. Dash curves represent the fitting curves. Cycle is the experimental data.
15154
J. Phys. Chem. B, Vol. 114, No. 46, 2010
The absorption of this peak is enhanced when the symmetry of the heme group distorts from D4h such that electron density at the metal and porphyrin changes.68 In addition, the crystal structure of ferric Cyt c investigated by X-ray and NMR methods shows that one of the water molecules is located inside the distal heme cavity and relatively close to the MET80-Fe bond.69,70 Therefore, when the film samples are hydrated, the water may extend to the Fe with the peak at ∼695 nm subsequently affected by the hydration of the protein. Recently, Zelent et al. studied the effects of the water on charge transfer (C-T) band of ferric heme using horseradish peroxidase as the modeling protein and confirmed that the C-T band is sensitive to the hydration of the protein.71 Our results suggest that the water molecules can affect the local electric fields and thus reduce the symmetry of the heme group, which will change the interactions of the iron d orbital with porphyrin π orbital, and affect MET80-Fe bond environment and the energy relaxation time. 3.2. Transient Absorption Measurements. Different probe wavelengths may influence the results in terms of what they interrogate in the dynamics of the protein. Recently, the Zhong group has systematically investigated the ultrafast dynamics in ferric Cyt c with the excitation at 400 nm and a wide range of probing wavelengths from the visible (700 nm) to the UV (280 nm).33 Here we have measured the hydration dependence of transient absorption probed at 400 and 800 nm after Soret band (400 nm) excitation. Because the 800 nm probe is far from Soret band, Q-band, and the 695 nm C-T band, we believe that there is no stimulated emission from 800 nm and the resulting relaxation time from two-color experiments may approach the pure intermolecular vibrational relaxation time of the ferric Cyt c. Representative results of the hydration dependence of transient dynamics probed at 400 and 800 nm are shown in Figure 6. Recent studies on the photoactivation of ferric Cyt c solution indicate that the dynamics mainly occur at the local site, including ultrafast internal conversion in hundreds of femtoseconds, vibrational cooling on the picosecond time scale, and complete ground-state recovery in 10 ps.33,37-39 Therefore, we only measured spectroscopic changes up to 8 ps. Two separate dynamic components after Soret photoexcitation were observed both in single-color and two-color measurements: a large amplitude fast component and small amplitude slow component. The transient dynamics can be well characterized with a multiexponential decay function
Ye and Markelz
Figure 6. (A) Hydration (g water/g protein) dependence of transient dynamics probed at 400 nm after Soret state excitation: (a) h ) 11.4%; (b) h ) 16.5%; (c) h ) 23.5%; (B) Hydration (g water/g protein) dependence of transient dynamics probed at 800 nm after Soret state excitation: (a) h ) 8.4%; (b) h ) 16.9%.
∆T ) H(t - t0){A exp(-(t - t0)/τ0) + B exp(-(t - t0)/τ1) + C exp(-(t - t0)/τ2) + D} (2) where H(t) is the Heaviside function that accounts for the pump-probe cross-correlation time; A, B, and C are constants correcting for amplitude. D is the constant offset and t0 is the time zero. τ0 represents the Soret state excitation time, τ1 and τ2 (τ1 < τ2) describe the fast and slow energy relaxation processes. To avoid confusion, the relaxation time (τ1, τ2) is marked as τ11, τ12 for single-color results and τ21, τ22 for twocolor results. Some examples of the fitting details are given in the Supporting Information. The fast time constants may have large error (60-150 fs) because of entanglement with the coherence coupling signal. The final value of τ11, τ12, τ21, and τ22 are the average results of at least four data sets from different samples. The error bars show the deviation in the averaged relaxation time.
Figure 7. Hydration dependence of relaxation time probed at 400 nm with pump beam intensity of 1.6 mW/mm2 and probe beam intensity of 0.6 mW/mm2.
3.2.1. Hydration Dependence of Energy Relaxation Time Probed at 400 nm. Hydration dependence of energy relaxation time probed at 400 nm with pump beam intensity of 1.6 mW/ mm2 and probe beam intensity of 0.6 mW/mm2 is shown in Figure 7. As a comparison, we have also measured the transient dynamics in solution. Time scales of solution with τ11 ) 350 ((60) fs and τ12 ) 3.7 ((0.5) ps are in excellent agreement with those observed in ferric Cyt c by Jimenez et al. with τ11 ) 200 fs and τ12 ) 3.7 ps and by Lowenich et al. with τ11 ) 300 fs and τ12 ) 4 ps.37,38 Our results also correspond well to those in ferric Cyt c with Q-band excitation, where time constants of
