10540
J. Phys. Chem. B 2006, 110, 10540-10549
Ultrafast Solvation Dynamics of Human Serum Albumin: Correlations with Conformational Transitions and Site-Selected Recognition Weihong Qiu,† Luyuan Zhang,† Oghaghare Okobiah, Yi Yang, Lijuan Wang, and Dongping Zhong* Departments of Physics, Chemistry, and Biochemistry, Programs of Biophysics, Chemical Physics, and Biochemistry, 191 West Woodruff AVenue, The Ohio State UniVersity, Columbus, Ohio 43210
Ahmed H. Zewail* Laboratory for Molecular Sciences, Noyes Laboratory of Chemical Physics, California Institute of Technology, Pasadena, California 91125 ReceiVed: October 19, 2005; In Final Form: January 17, 2006
Human serum albumin, the most abundant protein found in blood plasma, transports a great variety of ligands in the circulatory system and undergoes reversible conformational transitions over a wide range of pH values. We report here our systematic studies of solvation dynamics and local rigidity in these conformations using a single intrinsic tryptophan (W214) residue as a local molecular probe. With femtosecond resolution, we observed a robust bimodal distribution of time scales for all conformational isomers. The initial solvation occurs in several picoseconds, representing the local librational/rotational motions, followed by the dynamics, in the tens to hundreds of picoseconds, which result from the more bonded water in the tryptophan crevice. Under the physiological condition of neutral pH, we measured ∼100 ps for the decay of the solvation correlation function and observed a large wobbling motion at the binding site that is deeply buried in a crevice, revealing the softness of the binding pocket and the large plasticity of the native structure. At acidic pH, the albumin molecule transforms to an extended conformation with a large charge distribution at the surface, and a similar temporal behavior was observed. However, at the basic pH, the protein opens the crevice and tightens its globular structure, and we observed significantly faster dynamics, 25-45 ps. These changes in the solvation dynamics are correlated with the conformational transitions and related to their structural integrity.
I. Introduction Biomolecular recognition is mediated by water motions, and the dynamics of associated water directly determines local structural fluctuation of interacting partners.1-3 The time scales of these interactions reflect their flexibility and adaptability. For water at protein surfaces, our recent studies4-6 of hydration dynamics (tens of picoseconds) suggest that these mobile surface water layer molecules are constantly in exchange with outside bulk water. For trapped water in protein crevices or cavities, the dynamics becomes much slower and could extend to nanoseconds.7-9 These rigid water molecules are often hydrogenbonded to interior residues and become part of the structural integrity of many enzymes.7 Earlier we studied protein hydration with tryptophan being at the surface.5,6 The purpose of this work is to study local water motions in various environments, from a buried crevice to an exposed surface, using site-selected tryptophan but with different protein conformations. The aim is to understand the correlation between hydration dynamics and conformational transitions and then relate them to biological function. The structure and hydration of human serum albumin (HSA) are important for the transport of fatty acids and for binding a great variety of metabolites, drugs, and organic compounds in the circulatory system.10,11 It is a single polypeptide chain * Authors to whom correspondence should be addressed. E-mail:
[email protected];
[email protected]. † These authors contributed equally to this work.
consisting of 585 amino acids. Under physiological conditions (pH ≈ 7), HSA adopts a heart-shaped three-dimensional structure with three homologous domains I-III (Figure 1); each domain contains two subdomains A and B that consist of 4 and 6 R-helices, respectively.12,13 The X-ray structure shows that two halves of the albumin molecule form a 10-Å-wide, 12-Ådeep crevice that is filled with water molecules.12,13 The only single tryptophan residue W214 in the protein is located in the binding site IIA at the bottom of the crevice (Figure 1). With the change in pH, HSA undergoes reversible conformational isomerization, and pH-dependent isomers were first demonstrated by Luetscher14 in 1939 and later systematically classified by Forster15 in 1960. At neutral pH, the conformation of HSA is in its common physiological state, referred as a norm form (N); an abrupt transition occurs at a pH value of less than 4.3, changing the N form to the so-called fast migrating form (F), and when the pH is less than 2.7, a further transition takes place from the F form to the extended form (E). On the other side of the scale, when the pH value becomes above 8, the N conformation changes to the basic form (B), and at a pH above 10 the structure transforms to another aged form (A). These transitions are summarized below.10
conformation transitions E 79 8 F 79 8 N 79 8 B 79 8A 2.7 4.3 8 10 pH values The protein has been extensively studied in every aspect due to its physiological importance and its potential as a drug delivery vehicle10,11 and because different acidic and basic
10.1021/jp055989w CCC: $33.50 © 2006 American Chemical Society Published on Web 04/08/2006
Ultrafast Solvation Dynamics of HSA
J. Phys. Chem. B, Vol. 110, No. 21, 2006 10541
Figure 1. Left panel: X-ray crystallographic structure of human serum albumin13 at the neutral pH condition. The three constituting domains I, II, and III are in purple, green, and blue, respectively. Note a deep crevice between the two halves of the structure with W214 (in yellow) located at the bottom. Right panel: Local configuration around W214 with eight trapped water molecules within 7 Å and four potential tryptophan fluorescence quenching residues in close proximity of less than 5 Å.
isomers may have certain biological functions.16,17 Most studies have focused on binding interactions with various ligands and revealed that HSA is an assembly of squirmy and resilient components that frequently change in conformation through opening and closing of major crevices. With this “breathing” motion, HSA assimilates and releases a variety of substances during transportation in the circulatory system. Many complex structures have also been resolved at high X-ray resolution.12,18,19 The flexibility to adopt different conformations with independent segmental movements has also been shown through certain independent and sequential folding/unfolding studies of individual subdomains.20,21 The tryptophan W214 and extrinsic labeled dye molecules have been used as optical probes to study the conformational dynamics through changes in fluorescence lifetime, resonance energy transfer, and time-resolved anisotropy.22-26 The significant change in the lifetime of W214 for a series involving mutation in the binding site IIA clearly showed the local flexibility.27 Previously, we reported studies of binding interactions28 and hydration dynamics of protein unfolding,21 elucidating the rigid hydrophobic recognition and the different water mobility in three domains. In this study, we conducted a systematic investigation of the femtosecond-resolved fluorescence dynamics of tryptophan (W214) in HSA with the aim to understand the dynamics and structural flexibility under physiological conditions (N form) and to reveal the dynamical changes for conformational transitions occurring at different pH values. Because it is known that water is present (eight water molecules shown in Figure 1) around tryptophan in the crevice of the N conformation the dynamics should reflect the influence of their motions. The paper is organized as follows. Section II describes the experimental methodology including the femtosecond up-conversion laser setup and sample preparation. In section III, we present the femtosecond-resolved experimental results, and the discussion is given in section IV, focusing on the correlation of our results with conformational flexibility and transition. Finally, we conclude the work in section V. II. Experimental Section A. Femtosecond Up-Conversion Laser Setup. All of the femtosecond-resolved measurements were carried out using the femtosecond-resolved fluorescence up-conversion apparatus described elsewhere.29 Briefly, the femtosecond pulse after the
two-stage amplifier (Spitfire, Spectra-Physics) has a temporal width of 110 fs centered at 800 nm with a pulse energy of more than 2 mJ and a repetition rate of 1 kHz. Half of the laser energy was used to pump an optical parametric amplifier (OPA-800C, Spectra-Physics) to generate signal (1242 nm) and idler (2248 nm) beams. The latter was mixed with the residual fundamental beam (800 nm) in a 0.2-mm-thick barium borate crystal (BBO, type I) to generate a femtosecond pulse centered at 590 nm. This laser pulse, which was compressed through a pair of prisms with double paths to produce a temporal resolution of 60 fs, was frequency-doubled to generate the pump wavelength at 295 nm using another 0.2-mm-thick BBO crystal. The 295-nm excitation wavelength was chosen to minimize tyrosine residue absorption and dominantly excite the single W214 residue;30 the observed fluorescence essentially results from the excited W214 emission. The pump pulse energy was typically attenuated to ∼140 nJ prior to being focused into the motor-controlled rotating sample cell. The fluorescence emission was collected by a pair of parabolic mirrors and mixed with a gating pulse from another half of the fundamental beam (attenuated) in a 0.2-mm BBO crystal through a noncollinear configuration. The up-converted signal ranging from 218 to 292 nm was detected by a photomultiplier coupled with a double-grating monochromator. The instrument response time under the current noncollinear geometry is between 350 and 450 fs as determined from the up-conversion signal of Raman scattering of water around 326 nm. The pump beam polarization was set at the magic angle (54.7°) with respect to the acceptance axis (vertical) of the up-conversion crystal, and the polarization of the gating beam was set parallel to this axis through a half-wave plate. For fluorescence anisotropy measurements, the pump beam polarization was rotated to either parallel or perpendicular to the acceptance axis to obtain the parallel (I|) and perpendicular (I⊥) signal, respectively. These transients were used to construct the time-resolved anisotropy: r(t) ) (I| - I⊥)/(I| + 2I⊥). B. Materials and Sample Preparation. HSA (essentially free of fatty acids and globulin) was obtained from SigmaAldrich in lyophilized powder and used without further purification. All of the other chemical reagents were purchased from Fisher Scientific. Protein stock solutions of 0.5 mM HSA were prepared by dissolving the lyophilized powder in corresponding buffers to make various isomers (N form, 10 mM sodium
10542 J. Phys. Chem. B, Vol. 110, No. 21, 2006
Qiu et al.
phosphate, pH 7.0; F form, 10 mM sodium acetate, pH 4.1; E form, 10 mM sodium acetate/HCl, pH 2; B form, 10 mM Borax, pH 9.0; and A form, 10 mM Na2HPO4/NaOH, pH 11.0). For all femtosecond-resolved measurements, protein solutions were used without further dilution while sample solutions were diluted to ∼10 µM for the measurement of steady-state fluorescence spectra. Protein stock solution at pH 11.0 was incubated at 4 °C for more than 24 h before any spectroscopic measurements. C. Steady-State Fluorescence Characterization. The steadystate fluorescence spectra were taken using a SPEX FluoroMax-3 spectrometer. All of the protein samples were excited at 295 nm to ensure the dominant fluorescence emission from the single W214. At neutral pH, the emission spectrum of HSA has an intensity maximum at 338 nm, similar to the surface tryptophan emission,5,6 indicating a highly polar environment with a large amount of trapped water around W214 in the crevice, as also confirmed by the crystallographic structure13 (Figure 1). The emission peak is shifted to the blue at 332.6 nm when pH is reduced to 4.1, suggesting a less polar microenvironment surrounding W214 when the protein undergoes N f F conformational transition. At pH 2.0, though the protein structure expands to its maximum extent allowed by the internal disulfide linkages, the rearrangement of subdomain structures leads to a further blue shift of the emission maximum to 330 nm, indicating a more hydrophobic environment around W214 in conformational transition from F to E. In the slightly alkaline pH region (pH 9.0), HSA exhibits an emission spectrum with a peak at 336 nm, indicating a slight change of local polarity around W214 upon N f B transition. At pH 11.0, no significant further shift of emission maximum was observed for the A form while the spectrum width becomes slightly narrower compared with the emission from the B form, suggesting similar local polarity. D. Construction of the Solvation Correlation Function. The solvation correlation function was constructed following the method developed by Zhong and co-workers.31 In brief, all femtosecond-resolved fluorescence up-conversion transients are described by a sum of two terms of discrete exponential functions popul Iλ(t) ) Isolv (t) ) λ (t) + Iλ
∑i Ri e-t/τ + ∑j βj e-t/τ i
j
(II.1)
where the first term, with parameters Ri and τi, represents solvation contribution while the second term, with parameters βj and τj, represents the intrinsic lifetime (population) decay. The oVerall time-resolved emission spectra are constructed as follows
I(λ,t) )
Iss λ Iλ(t)
∑i
Riτi +
∑j
(II.2) βjτj
where the Iss λ denotes the steady-state emission intensity at wavelength λ. In a similar manner, the lifetime-associated timedependent emission spectrum is constructed according to
I
popul
(λ,t) )
popul (t) Iss λ Iλ
∑i
Riτi +
∑j
(II.3) βjτj
The resulting time-resolved emission spectra, a function of fluorescence intensity versus frequency in cm-1, are further fitted
with a log-normal function to deduce the oVerall emission maximum Vs(t) and the lifetime-associated emission maximum Vl(t). Typically, the solvation time (τi) is shorter than the lifetime (τj). The complete solvation time is defined as τsc when the difference between Vs(t) and Vl(t) reaches 0.5 cm-1. The corresponding emission peak is denoted as Vsc. Apparently, these two emission maxima would no longer differentiate themselves after the solvation contribution diminishes. It is also readily realized that the probe tryptophan has not yet reached its steadystate emission peak Vss and the corresponding time (τss), when the femtosecond-resolved spectrum reaches its steady-state emission maximum, is longer than the complete solvation time τsc. After solvation is completed, the femtosecond-resolved emission spectra continue to move to the red side due to the existence of multiple lifetimes with distinct emission peaks. Thus, the solvation response function is constructed by subtracting the lifetime-associated emission contribution Vl(t) from the oVerall femtosecond-resolved emission maximum Vs(t) as follows
c(t) )
Vs(t) - Vl(t) Vs(0) - Vl(0)
(II.4)
The derived c(t) function is fitted with a multiple-exponential decay function. III. Results A. Femtosecond-Resolved Fluorescence Dynamics. A series of femtosecond-resolved fluorescence up-conversion transients of HSA for different isomers are shown in Figures 2-6. More than eight transients of emission wavelengths, spanning from the blue to the red side, were collected for each isomer configuration. For each conformation, all fluorescence transients were analyzed with a global fitting strategy. Typically, four discrete exponential functions were used, two of which are the lifetime emission decays. The two fluorescence lifetimes varied depending on pH value but were fixed for all transients of a given isomer. The lifetimes were determined carefully because they would affect the time scale of slow solvation dynamics. A series of transients at the red side of wavelengths, from 360 to 380 nm, was obtained, and the two lifetimes, with different ratios, are well fitted for these red-side transients. The longer lifetime is determined by conventional photon-counting measurement (from the literature), and the shorter one was obtained by fitting our 3-ns-long transients. Finally, these two lifetimes were used to fit all other transients, and all parameters from the fits are selfconsistent (Supporting Information). The longer lifetime of the N form is 6.1 ns,22 and the shorter one from our 3-ns transients was determined to be 1.4 ns. The red-side transients of the F form are nearly the same as those from the N form; thus the F form has the same two lifetimes. By comparison of red-side transients of other isomers, we determined the lifetimes of 912 ps and 4.6 ns for the E form, 1.15 and 6.1 ns for the B form, and 318 ps and 3.8 ns for the A form. However, it should be pointed out that because the time interval between two consecutive pump excitations is 2 ms and each transient was averaged over more than 1 h, for each isomer all fluctuated configurations would probably be sampled on such long times. Thus, the apparent two lifetimes do not mean that only two static configurations are present, but they mainly represent two types of temporal configuration distributions; the system is in dynamic heterogeneity.
Ultrafast Solvation Dynamics of HSA
Figure 2. Normalized, femtosecond-resolved fluorescence transients of the N isomer in the short (left) and long (right) time ranges, with a series of gated fluorescence emissions. Note that solvation dynamics is significantly slower than that of tryptophan in a similar buffer.31
1. N Form. Figure 2 shows the femtosecond-resolved fluorescence transients of W214 for several typical wavelengths. The overall decay dynamics is significantly slower than that of aqueous tryptophan in a similar buffer solution.31 Clearly, the ultrafast decay components (
